How to treat an external charger?

[g.didom]

Hello,
we are trying to certify a Class I medical device. The device is internally powered, i.e., it contains a Li-ion battery which supplies power during usage.
We are struggling to understand how to treat the associated charger, which is an external component which converts mains to a 19V. I know this question has been asked multiple times, but even the answers are very application-specific and sometimes contradictory.

The charger is shipped toghether with our core product and supplied off-the-shelf, and by itself is NOT 60601 compliant; however, it has all the standard certification for a quality non-medical charger and comes from a reputable manufacturer. It is a traditional charger that goes into mains, has a transformer, and then another cable which supplies the voltage to the DC port of our device.

The charger is not meant to be connected while the device is used, as functionality is blocked; however, it is possible (foreseeable misuse) for the patient to wear the device while keeping the charger connected, despite no functionality is present and clear indications to avoid doing so have been provided.

We are evaluating the following options:

• Treat the charger of part of our ME EQUIPMENT.
  • Which are the implications? Do we need to re-label and claim responsibility for the charger? For instance, do we need to include the materials of the charger in our biocompatibility assessment?
  • Do we need to perform tests for MAINS PART connection in this case?
  • Do we need to account for the charger rating when supplying the rating of our device?
• Treat the charger as part of an ME SYSTEM made of our ME EQUIPMENT device and the charger.
  • Does this opens us to any additional testing arising from 60601-1 Clause 16?
  • Does the carger classify as a power supply, even if the power is sourced from the battery during usage?
• Treat the charger as an ACCESSORY. this is what was suggested from our test lab; however, we are not convinced of this narrative as the charger only serves for recharging, i.e., does not help the device achieve its INDENDED (medical) USE.

The definition of ACCESSORY to me is quite contradicting. From Terminology and definitions of IEC 60601-1:

ACCESSORY
additional part for use with equipment in order to:
– achieve the INTENDED USE,
– adapt it to some special use,
– facilitate its use,
– enhance its performance, or
– enable its functions to be integrated with those of other equipment

So, it suggests ACCESSORIES need to be considered as such only when the medical (INTENDED) purpose is affected by the employment of the accessory; however, in NOTE 1 of 3.63 MEDICAL ELECTRICAL EQUIPMENT definition:

NOTE 1 ME EQUIPMENT includes those ACCESSORIES as defined by the MANUFACTURER that are necessary to enable the NORMAL USE of the ME EQUIPMENT.

So, it refers to NORMAL USE, i.e., also includes maintenance, transport, ... and is in general better applicable to a charger which does not have any medical purpose nor certification.

Note in Subclause 7.9.2.14 is even more explicit, but does not mention ACCESSORIES at all:

NOTE What was referred to in the first and second editions of this standard as a “specified power supply” is considered either as another part of the same ME EQUIPMENT or as other equipment in an ME SYSTEM. Similarly, a battery charger is considered either as part of the ME EQUIPMENT or as other equipment in an ME SYSTEM.

To add a bit more of entropy, we have performed some external lab tests in which the charger was shipped with our product; some tests (i.e., leakage currents) have been performed taking into account the charger. Some others (i.e., IP22, drop tests, ...) did NOT take into account the charger, nor the device+charger (connected) system.

 

[d_addams]

Hello,
we are trying to certify a Class I medical device. The device is internally powered, i.e., it contains a Li-ion battery which supplies power during usage.
We are struggling to understand how to treat the associated charger, which is an external component which converts mains to a 19V. I know this question has been asked multiple times, but even the answers are very application-specific and sometimes contradictory.

The charger is shipped toghether with our core product and supplied off-the-shelf, and by itself is NOT 60601 compliant; however, it has all the standard certification for a quality non-medical charger and comes from a reputable manufacturer. It is a traditional charger that goes into mains, has a transformer, and then another cable which supplies the voltage to the DC port of our device.

The charger is not meant to be connected while the device is used, as functionality is blocked; however, it is possible (foreseeable misuse) for the patient to wear the device while keeping the charger connected, despite no functionality is present and clear indications to avoid doing so have been provided.

We are evaluating the following options:

• Treat the charger of part of our ME EQUIPMENT.
  • Which are the implications? Do we need to re-label and claim responsibility for the charger? For instance, do we need to include the materials of the charger in our biocompatibility assessment?
  • Do we need to perform tests for MAINS PART connection in this case?
  • Do we need to account for the charger rating when supplying the rating of our device?
• Treat the charger as part of an ME SYSTEM made of our ME EQUIPMENT device and the charger.
  • Does this opens us to any additional testing arising from 60601-1 Clause 16?
  • Does the carger classify as a power supply, even if the power is sourced from the battery during usage?
• Treat the charger as an ACCESSORY. this is what was suggested from our test lab; however, we are not convinced of this narrative as the charger only serves for recharging, i.e., does not help the device achieve its INDENDED (medical) USE.

The definition of ACCESSORY to me is quite contradicting. From Terminology and definitions of IEC 60601-1:

So, it suggests ACCESSORIES need to be considered as such only when the medical (INTENDED) purpose is affected by the employment of the accessory; however, in NOTE 1 of 3.63 MEDICAL ELECTRICAL EQUIPMENT definition:

So, it refers to NORMAL USE, i.e., also includes maintenance, transport, ... and is in general better applicable to a charger which does not have any medical purpose nor certification.

Note in Subclause 7.9.2.14 is even more explicit, but does not mention ACCESSORIES at all:


To add a bit more of entropy, we have performed some external lab tests in which the charger was shipped with our product; some tests (i.e., leakage currents) have been performed taking into account the charger. Some others (i.e., IP22, drop tests, ...) did NOT take into account the charger, nor the device+charger (connected) system.

a charger clearly contributes to achieving intended use (without the charger the device will no longer function once the primary cell depletes in relatively short order compared to the intended life of the device). So this is not contradictory. From a system perspective the charger does not have to be the 'business' end of device providing 'the last mile of therapy'.

If the product becomes unusable if a component malfunctions, that component can be considered contributing to intended use. Yes a charger itself has no medical function, but you aren't distributing a charger for a variety of uses, you are providing it as part of a medical device intended to deliver some benefit and without the charger the device cannot achieve its stated mission. Don't try to get cute with rationales like "well the device can last a day before needing recharging so the charger isn't necessary". The basic test for this are questions like: do we only sell a version without a charger? If the charger breaks would customers replace it? If the answers to these are 'yes', the charger seems to be necessary to achieve intended use.

 

[Peter Selvey]

This is a major area where IEC 60601-1 is out of step with regulation, law and common sense.

According to the law, it is OK for a manufacturer of a medical "main unit" to create a medical system using things from other manufacturers, some which could be medical (e.g. custom power supplies, medical sensors, electrodes, tubing, needles etc.), and some not (e.g. general purpose power supplies, cables for ac, USB etc.). There is no requirement for the main unit medical manufacturer to take over responsibility for those other objects as a "manufacturer" for those objects.

Moreover, it's also OK for users to create their own medical system, e.g. using accessories that are not explicitly specified by the manufacturer of the "main unit", but are nevertheless are compatible. A examples are electrosurgery systems, patient monitors, infusion/syringe pumps which in the real world are often used with accessories manufactured by companies that are completely independent from the manufacturer of the main unit. In the less serious world, it could be a main unit with a Type C USB connector, that can be charged by any charger with a Type C output. It sounds messy and uncontrolled, but it works, and it's actually beneficial to do it this way, rather than trying to have a single manufacturer responsible for the whole system.

It turns out there are many logistical issues with trying to assign one manufacturer as being responsible for a system, or trying to have a single TYPE TEST report cover a system, especially where there are several manufacturers involved. I could write a five page article on this. In general, regulations are carefully designed to apply only to individual medical devices. There are of course many issues raised by systems (such as thinking about device combinations, compatibility, risk controls) but regulations are careful not to assign one manufacturer responsible for the system, see MDR Article 22 as an example. Each device manufacturer should think about the use in systems carefully, but the results of that analysis should only result in design and production controls for their device, and only these aspects should be considered in a TYPE TEST.

The real kicker is that attempts to combine more than one device in a single TYPE TEST report can and does actually lead to reduced safety, which is counterintuitive but true. Consider an example where the accuracy of a system comprised of a main unit and a sensor (each independently manufactured) is required to be ±10%. Lets say both manufacturers stupidly use this criteria in their design, which in the market could result in systems up ±20%. However a TYPE TEST on a single sample of a "system" could easily pass. One unit could be -6% the other +9% and together results in +3%, a "Pass". Don't laugh, stuff like this happens. In fact, statistically it's more likely that system testing will hide problems that would have been picked up by individual device testing against a reasonable individual device criteria derived from the system (e.g. ±5% for each device, which would have failed each device in the above example). And the real critical point here is not the TYPE TEST, but that each manufacturer uses the correct criteria (e.g. ±5%) in design and production, something that is not ensured by focusing on the system criteria (±10%).

Within IEC 60601-1, the following clauses are in contradiction:

• 3.63, Note 1 states that ME EQUIPMENT includes accessories
• 3.55 states that the manufacturer of the ME EQUIPMENT is responsible for the design, manufacturer, labelling, packaging of the ME EQUIPMENT (and hence, combined with 3.63 makes them responsible for the accessory, not possible)
• 4.2.2 (risk management) states that the term "medical device" is replaced with ME EQUIPMENT, and thus requires ISO 14971 to formally cover accessories of other manufacturer, not in a "device compatibility" view but full 100% responsibility for for the accessory (again not possible)
• 7.2.2 (accessories) and 7.2.5 uses phrases indicating there can be other MANUFACTURERs involved, which is not possible legally (a basic principle of declaration of conformity, TYPE TEST and regulations is identifying a single manufacturer that is responsible for design and production, there cannot be multiple manufacturers involved)

At some point, all of these contradictions (and others) will have to get resolved, and the standard aligned with common sense and regulations.

For now, the big problem is test labs, auditors and regulators that might take IEC 60601-1 at face value and insist on reports that cover systems. Generally, this is handled by a fudge, test labs kind of close their eyes and make the reports look as if they cover the system. I think as long as everyone is aware it's a fudge and makes sure that the real "device" is OK, it can work.

One point of confusion is that in order to test an individual device, it can be necessary to set up a system with other devices. For example, in the original question, you might want to run a temperature rise test during charging mode to make sure components in the main unit don't get to hot, especially the battery. Obviously for this, you will need to use an external charger. But this itself doesn't mean the charger comes under the scope of test, it's just one of the objects used to support the test. And, let's say the analysis indicates that the charger's voltage could affect the main unit's temperature, so testing with a random charger might not get the worst case result. In that case, the best approach could be to use the specifications of the charger (e.g. 19V ± 10%), and then replace the charger with a variable dc supply and test at 17V, 19V and 21V to be sure. That's an example of how taking into account the system affects the TYPE TEST, without the TYPE TEST covering the system.