Is IEC 60601-2-18:2009 compatible with IEC 60601-1 ed. 3.2?


My question is IEC 60601-2-18:2009 compatible with IEC 60601-1 ed. 3.2?
Is it possible to test in accordance with IEC 60601-2-18:2009 and at the same IEC 60601-1 ed. 3.2?


Involved In Discussions
If you pay the test lab enough, they will test anything you want ;-)
I would also test for the most recent norm in all cases. Just be aware that the particular norm overwrites the standard norm.

Peter Selvey

Super Moderator
Officially no. A particular has a dated reference to the general standard and this should remain accurate.

The common practice of listing both the "General Standard" and the "Particular standard" as separate line items in a declaration of conformity or in a test report is not strictly correct and is the root cause of misunderstandings. Only the particular should be listed and it essentially forms a single, stand-alone standard that incorporates a modified (mashed up) general standard as instructed by the particular. Logically (logistically?) it must be a dated reference to the general standard to allow the standard committee to review the updates to the general standard and make sure they are compatible. For example, the particular might say "delete the third paragraph" while an amendment to the general adds a new first paragraph, misaligning the instruction by the particular, and resulting the reader deleting the wrong paragraph.

Because of this, particulars must always have a dated reference to the general, it's not possible to have an undated flexible, play it by ear reference. IEC 60601-2-18:2009 makes a specific reference to IEC 60601-1:2005 (Ed 3.0) and says nothing about the amendments or new editions.

Unfortunately, the common practice of listing both the particular and the general raises the problem of less experienced people seeing the older version of the general and asking questions (auditors, technical file reviewers etc.). Also, this gives committees for particular standards less incentive to update their standards in a timely manner, as is obviously the case here. The 2009 is a long time ago ...


Involved In Discussions
How can this approach be reconciled with the requirement that the medical device is state of the art if it is not tested according to the current state of the art (the current version of the main standard)?

Peter Selvey

Super Moderator
Because many 601-2-xx standard committees have been terribly slow and the EU harmonization process even slower (virtually dead?), many Notified Bodies are leaning heavily on "state of the art" as a way push standards.

I understand why they do this, but it is actually a misuse of "state of the art".

A normal standards update process might involve:

- general standard is amended and published
- regulations provide a 3 year transition period, but this only applies to devices that have no particulars
- related standards in the series (particular, collaterals) should review the amendment and update their standards in a timely manner (e.g. 1-2 years)
- the updated related standard (particular, collateral) is then published
- regulations give a 3 year transition period for those standards (i.e. 4-5 years after the general standard)

If this time schedule is followed, then using the "state of the art" to push standards ahead of schedule is just plain wrong, it's illegal and irresponsible. If there is a serious situation that requires quicker update, other mechanisms should be used, for example in Europe there is MDR Article 98 which allows quick action outside of standards.

The problem here is that we have a standard that has not been updated for 14 years and hence is now well out of date not only with the general standard, but also EMC and other particulars it refers to (for example, IEC 60601-2-18:2009 has dated references to IEC 60601-1:2005, IEC 60601-1-2:2007 and IEC 60601-2-2:2009, all of which are considerably out of date). This is unlikely to be a serious safety issue but it does make it difficult for test labs, CB scheme, manufacturers, regulators and everyone involved, since in general most agencies have moved on to the latest standards by now.

So I get why a workaround might be necessary here, and it may be necessary to mash up the old and new standard with a little common sense thrown in. But care should be taken to avoid applying this generally or as a rule. This is a special problem due to an excessive delay in updating the particular standard. If the standards committee is working in a timely manner, the normal 4-5 years should be allowed according to the rules without anybody throwing around "state of the art".
Top Bottom