How to Classify ME Equipment and Identify Applied Parts

Nothing downstream can be scheduled until the classification is settled

When an active medical device reaches the laboratory, the first thing the engineer does is not to connect a test instrument. It is to open the accompanying documents and the electrical schematics and answer three questions. Does this device fall within the scope of medical electrical (ME) equipment? By what means does it achieve protection against electric shock? Does it have applied parts, where exactly are they, and what type is each of them?

Until those answers exist, there is no basis for deciding how the leakage current measurement is wired, between which two points the dielectric strength test voltage is applied, along which segment creepage distance is measured, or where electromagnetic disturbance is injected. The whole body of requirements in GB 9706.1 and IEC 60601-1 is conditional: for one and the same product, a different classification produces a completely different set of applicable clauses.

In real projects, rework is rarely caused by a badly built product. It is caused by getting this step wrong. The class marked on the rating plate does not match the internal construction; an applied part is overlooked; the status of an external adapter is never stated; a battery-powered product is forced into a category that does not actually apply to it. When problems like these surface during testing, they usually mean the construction has to change, not that a setting can be tweaked. This article takes the step apart so that you can get the positioning right before samples ship.

Question one: is it ME equipment at all

The criterion is not "it is used in a hospital". It is whether several characteristics hold at the same time: there is a physical or electrical connection to the patient, or energy is transferred to or taken from the patient; the equipment is supplied from the supply mains or from an internal electrical power source; and the intended purpose is diagnosis, treatment or monitoring. For the exact wording of these characteristics and how they combine, refer to the definitions in the current valid versions of GB 9706.1 and IEC 60601-1.

Three grey areas come up repeatedly in engineering practice.

A mechanical product with a motorised mechanism bolted on. Adding a motor does not automatically make a product ME equipment. What matters is whether the electrical portion participates in delivering the diagnostic or therapeutic function, and whether it forms a connection to the patient.

A terminal that only displays or processes information and does not itself acquire signals from the patient. Whether such a product is in scope depends on its position in the overall clinical chain and on the manufacturer's declared intended use.

A system assembled from ME equipment and non-ME equipment. Here the object of assessment is the system, not the single unit, and the evaluation boundary widens noticeably.

One documentation gap causes rework constantly: printers, displays, external power supplies and wireless modules arrive together with the sample, but nothing states whether they belong to a combination specified by the manufacturer. The laboratory has to fix the boundary first, because once the boundary expands, the test points, the documents required and even the number of samples all change with it. Before submission, put "device proper + accessories specified by the manufacturer + optional configurations" on a single sheet and send it with the samples; for the expected document format, see the sample submission requirements.

Question two: how is shock protection achieved - Class I, Class II, or internally powered

By means of protection against electric shock, the usual positioning has three parallel branches rather than two options: Class I, which relies on protective earth; Class II, which relies on double or reinforced insulation; and equipment supplied from an internal electrical power source.

The third branch is often treated as a footnote to the first two. In fact it is independent. Handheld, portable and wearable active products that run on batteries mostly land there. If the form you are filling in pushes you into a false choice between "Class I or Class II", the starting point is already wrong, and the isolation diagram, the leakage current wiring and the dielectric strength test points will all drift along with it. That rework costs far more than asking one extra question up front.

The evidence for Class I is not an earth symbol printed on the enclosure. It is whether accessible conductive parts are genuinely connected to the protective earth terminal of the power inlet through a continuous protective earth path, and whether the way that path is made satisfies the reliability requirements of the standard. The core of Class II is that safety does not depend on the earth conductor provided by the building; the insulation construction constitutes complete protection on its own.

Internally powered equipment needs one more layer of thinking. The key question is whether it stays isolated from the supply mains throughout its whole use cycle.

• Equipment that is never connected to the supply mains in any state and runs only on its internal source is assessed as internally powered equipment.
• As soon as it can be connected to the supply mains for charging, that state is no longer "purely internally powered". How shock protection is achieved in that state, whether by protective earth or by double or reinforced insulation, decides which class applies in that state.
• Equipment that supports use while charging needs particularly clear description, because a patient connection and a mains connection exist at the same time; declaring only one of the two states is not acceptable.
• Whether the charger or power adapter is part of the equipment must be declared explicitly by the manufacturer, because the declaration itself changes the evaluation boundary.

Filling this table in before submission usually removes most of the pre-test clarification loop. How many supply states the device has, how shock protection is achieved in each of them, and whether a patient connection exists in each: with those three columns complete, the laboratory can draft the test plan directly.

What follows is engineering analysis of construction and current paths, not a statistical conclusion. Seen from the energy path of shock protection, Class I equipment places part of its safety on an external condition, namely that the earth at the hospital socket is effective. Class II builds protection entirely into the insulation construction. Internally powered equipment cuts the mains path away completely while running on its battery, at the price of having to explain the charging segment separately. So, in the same use scenario, choosing Class II means the design pressure on insulating materials, creepage distances and clearances is carried entirely in-house; choosing Class I looks easier but brings a whole additional body of verification around protective earth impedance, the fixing and anti-loosening of earth conductors, and the continuity of accessible metal parts; going the internally powered route saves part of the mains-side requirements and adds the work of defining states and justifying the charging path. That trade-off belongs in the structural design phase, not in the week you discover that neither end is satisfied.

Contradictions that turn up in construction:

• The rating plate says Class II, but a functional earth was added to pass EMC, or the enclosure was tied to the mid-point of the filter capacitors to push leakage current down. Structurally the declaration no longer holds.
• An external power adapter is used, but nothing states whether the adapter is part of the equipment. This single point decides whether the adapter is pulled inside the test boundary or treated as an external supply.
• The device has several supply inputs, for example mains, battery and vehicle power, and the class in each state is not stated separately.
• The rating plate marks the device as internally powered, yet the unit carries a charging port that connects to the mains, and the instructions say nothing about how the charging state is classified.

Question three: which parts count as applied parts

"Whatever touches the patient is an applied part" is a widespread misreading. The actual assessment looks at two things: whether physical contact with the patient is necessary in normal use for the equipment to perform its intended function, or whether the manufacturer assigns the part to the patient circuit. Incidental, non-functional contact does not by itself create an applied part.

Three neighbouring concepts have to be kept apart.

One trap recurs: a device has several applied parts, possibly of different types, but the manufacturer declares and tests only one of them. A device may have both adhesive electrodes and a handle that the patient has to grip; the nature of the contact and the conductive path are entirely different, so marking, isolation requirements and testing all have to be handled separately.

Choosing Type B, Type BF or Type CF

Choosing the type of an applied part is essentially a combined judgement on three dimensions: how close the contact with the patient is, whether a low-impedance conductive path is formed, and whether the part may be used in direct cardiac application. "F-type" means that the applied part is isolated, or floating, from other parts of the equipment, which limits the possibility of external potentials reaching the patient.

This section deliberately gives no numerical patient leakage current limits, and the reason is worth stating. The values are not "one type, one number". Within a single type, normal condition and single fault condition are two separate sets of requirements; d.c. and a.c. are assessed separately; whether mains voltage is applied to the applied part and whether the part is defibrillation-proof also change which row applies. Compressing all those conditions into one memorable table is precisely where self-assessment goes wrong. For the isolation means, the number of means of protection and every limit value, check clause by clause against the current valid versions of the GB 9706.1 and IEC 60601-1 standard texts, and copy each value together with the conditions it applies to, never merging values across conditions.

Three common mistakes when choosing a type:

"CF is the strictest, so I will declare everything CF." The type is not a slogan that can be tightened at will. Once declared, the corresponding isolation has to be genuinely implemented in the construction and stated consistently on the rating plate, in the instructions for use and in the accompanying documents. Declaring what the construction does not deliver is itself a non-conformity.

Treating "defibrillation-proof" as part of the type. Defibrillation protection is a separate characteristic layered on top of the applied part type. It has to be declared and verified separately, and it is never included by default.

Marking the type only in the instructions. If the type is written in the instructions but not marked on the equipment as required, or the symbol marked differs from what the instructions say, the report records a non-conformity just the same.

How the outcome propagates: the isolation diagram is the key deliverable

Once the class and the applied parts are settled, the next deliverable is the isolation path diagram. It lists every insulation path leading from live parts to the applied parts, to accessible parts and to signal input/output parts, notes what implements each path (solid insulation, spacing, protective earth, current-limiting impedance), and states whether the path provides operator protection or patient protection. Internally powered equipment needs one extra path drawn for the charging state, because that path exists only while charging and is often the weaker segment in the overall insulation coordination.

Once the laboratory has that diagram, the dielectric strength test points, the locations for measuring creepage distances and clearances, and the leakage current wiring are essentially fixed. Submissions without an isolation diagram tend to spend more effort on clarification than on the testing itself, something that shows very clearly during intake for electrical safety testing.

The same set of decisions propagates into electromagnetic compatibility: how applied parts and patient connections are identified determines directly how the patient coupling port is defined in EMC testing. Both submissions therefore have to use one and the same classification, and inconsistent wording is exactly the kind of contradiction that gets picked up during technical review. Port definition and test arrangement belong to the EMC side; see EMC testing services.

What to prepare for submission

One issue around model families is routinely underestimated. A registration unit often contains several models, and if they differ in supply arrangement, in the number of applied parts or in applied part type, you cannot simply pick one model to cover the rest. The coverage argument has to be made on whichever dimension the difference appears, and additional testing may be required. Differences in battery model and power adapter model are especially easy to overlook: the same main unit with a different adapter may change the means of shock protection entirely, and in that case a coverage argument does not hold. Putting the difference table together before submission is usually less work than adding samples mid-project; for the workflow, see the testing process.

Back to the starting point

Classifying the equipment and identifying the applied parts looks like filling in a few form fields. In practice it is a complete retelling of the product's shock protection logic. If the story holds together, the testing that follows is verification. If it does not, testing turns into a health check on the design.

SUNGO Lab operates laboratories in Shanghai and Hefei and is accredited by CNAS, CMA and IAS (USA). We handle electrical safety and electromagnetic compatibility testing for medical electrical equipment, and we can help you work through equipment classification, applied part determination and the isolation path diagram before samples are shipped. Please note that an accreditation mark only demonstrates that the laboratory is technically competent within its accredited scope; it does not constitute a commitment regarding market access outcomes. Our full service scope is listed under testing services.

If your product is stuck at this step, call +86 132 4819 8029 or request a quote. Send us the schematics and the intended use statement, and we will sort out the positioning first and discuss the test plan afterwards.