Harmonization date of IEC 60601-1 , 3rd Edition

everybody, EN60601-1 2nd is not listed in recent OJ from 2012-8-30. Only 3rd is list. But some particular standards are not harmonized with 3rd via OJ. For products applicable to these particular standards, which general standard should we use? Can we still rely on "Frequently asked Questions related to the Implementation of EN 60601-1:2006 with respect to MDD 93/42/EEC? issued by 60601-1 Issues Team of NB orgnization on Jan 24, 2012 ? According to the guideline, particular standard decide transition period of 3rd.

Yes, it's OK to keep using the 2nd edition if the particular still refers to it.

Technically, you should be declaring compliance with only the particular standard. The IEC particulars are written in a way such that the two standards mesh together to create in effect a single standard, under the title of the particular (e.g. EN 60601-2-xx). In the same way, general and particular standards are not intended to be evaluated separately: there should really be only a single test report with the meshed requirements.

So as long as the old particular remains harmonized, the general standard also remains harmonized, through this meshing process.

Good point. Thanks Peter.
Our NB requested to give compliance evidence for 3rd general standard ,although the particular standard is yet not harmonized. It's very weird.

Peter Selvey said:

Yes, it's OK to keep using the 2nd edition if the particular still refers to it.

Technically, you should be declaring compliance with only the particular standard. The IEC particulars are written in a way such that the two standards mesh together to create in effect a single standard, under the title of the particular (e.g. EN 60601-2-xx). In the same way, general and particular standards are not intended to be evaluated separately: there should really be only a single test report with the meshed requirements.

So as long as the old particular remains harmonized, the general standard also remains harmonized, through this meshing process.

Click to expand...

The request from the NB strongly suggests an auditor not really familiar with how the 601 series works and is just running on a high level checklist. It can get complicated and the auditors have to cover a lot of ground, so just gently ask them to double check.

The NB-MED Frequently Asked Questions document on the 3rd edition was ordered to be withdrawn by the Directorate General of the European Commission, since their document indicated that "state of the art", the legal requirement in paragraph 2, annex 1 (Essential Requirements) of 93/42/EEC could not be met, unless newer harmonized standards were applied, or at least re-evaluated against. The Notified Bodies nor NB-MED have the authority to issue this kind of guidance on the European Directive.

Although some foreign websites (outside Europe) still have copies of the NB-MED document, it is no longer valid. Furthermore, Notified Bodies are not permitted to interpret "State of the art" in such narrow terms that a manufacturer must use voluntary [harmonized] standards. The NB-MED Q&A document indirectly implied that standards drawn up by ?private law bodies? (IEC, ISO, ASTM, AAMI) created new mandatory ?Public Law? criteria. This narrow interpretation is not acceptable and now the European Commission is even contacting individual notified bodies and telling them to remove such inferences from their website ?IEC 60601-1 3rd edition mandatory in Europe?.

A manufacturer can meet the Essential Requirements using any methods it chooses, so long as those methods reasonably address the ER. It should be noted that it is now well known that EN 60601-1:2006 grossly overstated how much of the ?Essential Requirements? it actually met in its Annex Z. So those that had based their CE mark on using 3rd edition will need to revisit their technical files.

Harmonized standards were intended to help, and they certainly do. Each manufacturer has to find the methods that work best for them, and hope they are making a convincing case to their notified body. There has been a lot of misunderstanding about the role of standards in conformity assessment, even among notified bodies. This doesn?t have to remain so uncertain.

The GHTF "Role of Standards in Conformity Assessment" was developed to demonstrate conformity with GHTF Essential Principles. It describes a list of acceptable methods for meeting the GHTF EPs. More importantly, it indicates that legacy products you still sell, that are known to be safe and effective (based on post market data) do not need to be re-assessed to newer standards. If you look at the Australian Essential Principles Checklist, you will also see clear evidence that they adopt the GHTF approach that accepts methods and standards, even standards that are of another origin.

The Director of Standards at the US FDA explained that; Any [appropriate]methods that support the manufacturers claim of safety and performance will be accepted, regardless of whether the methods were borrowed from older, even withdrawn standards". The "oldies but goodies" viewpoint sustains a manufacturers elective choice to use ?obsolete? standards for FDA 510(k) submissions. The FDA continues to accept the GHTF Essential Principles Checklist their ongoing 510(k) Pilot Program. This follows the GHTF view that ?methods? do not demand an origin with IEC, ISO, AAMI, etc, etc. SDOs), It is all about the relevance of the "method" to meeting the Essential Principles. Private law standards are considered a collection of technical methods that each manufacturer can elect to use, or not use, in part, or in whole, to meet the REGULATORY Public Law criteria ?Essential Principles/Requirements..

The only exception to this is OSHA, which demands that the manufacturer meet UL 60601-1 First Edition (US deviation of IEC 60601-1 2nd edition). OSHA will not currently recognize 3rd edition for meeting 29 CFR 1910 relating to the electrical safety of electrical equipment. In effect, you can use your UL 60601-1 approval or OSHA and FDA, but you cannot use your 3rd edition certification to meet OSHA electrical safety requirements.

This is very good information but I would provide a couple of caveats and extra info:

It is actually ISO 14971 that promotes the use of standards as state of the art (see the clause on policy for acceptable risk, and also rationale). Of course, you could argue that ISO 14971 itself is just a harmonized standard and not law, but the key point is that this is more than just a wild interpretation by the NB's, rather it is a well established internationally accepted approach.

Also, you cannot just ignore a harmonized standard. You need to know where you stand and then provide an alternate solution. So you still need a clause by clause check of a harmonized standard, it's just that an individual "fail" or "not evaluated" result is OK if justified. You can't just say I don't like EN 60601-1:2006 so I will just ignore it.

There are a lot of other problems with the NB-MED document such as the decision to informally allow 3 years for a brand new standard. There is no legal basis for this, and many other questionable interpretations in the document, so it's no surprise the EU came down hard on them.

But the real problem here is that the 3rd edition was so poorly written. In fact, a lot of IEC based standards are poorly written. Having worked with other standards such as AAMI, JIS, ISO it's quite clear that there is bias towards overkill in the IEC, a poor understanding of risk, and worse a lack of any attempt to validate of the standards, to make sure they work in practice. The alarm standard IEC 60601-1-8 is a classic case that is full of both overkill and technical errors which make it impossible to apply.

We should really focus on the core problem: poor standard writing, not the EU's system of harmonized standards.

Peter Selvey said:

We should really focus on the core problem: poor standard writing, not the EU's system of harmonized standards.

Click to expand...

Of course, standards are written by people, mostly volunteers as I understand it...who may be busy, likely are imperfect, may have multiple responsibilities leading to conflicting mixes of motivations, may have differing levels of technical knowledge/experience and wisdom, and so forth.

It's important to note that "poor standard writing" sometimes merely reflects that various stakeholders have different views of what a "good" standard would do.

Europe's shift from the Directive approach to a Regulation, more like FDA's approach, may help.

Therin lies the most critical difference between standards drawn up by Private Law Bodies and a regulation that is vetted through rigorous public commenting and legislative processes. There are Legistative and Executive Orders that new regulations have to be vetted through as well that demand cost/benefit, environmental impact, etc assessment. IEC, ISO, AAMI standards don't have to go through any of this.

One reason OSHA cannot so easily recognize a more expensive standards; There are no less than two Presidential Executive Orders that prevent regulations from being implemented, without a cost impact analysis to industry. As evidence reviewed by OSHA established that 60601-1 3rd edition was going to be more expensive than 2nd edition, and didn't provide any better protections for electrical safety, replacement of the less expensive standard with the more expensive standard could not be justified as meeting the criteria of the Executive Orders. Since UL 60601-1 had been tied inseparably to compliance with 29 CFR 1910.303 concerning electrical safety of medical devices, OSHA had to examine the cost/benefit issues of the 3rd edition. This occured simultaneously with a 2010 Executive Order intended to protect industry from costly regulatory changes during our economic crisis.

It is mind boggling to me that government is doing a better job protecting industry from costly standards, than the industry reps that are writing these standards. In all the Q&A etc you ever read about 60601-1 3rd edition, did the testing labs ever answer the question; How much more expensive is it? Where is the value in the extra cost? Everyone screams about "cost" except when it comes to standards? This may be why more regulations will likely avoid being tied so closely to standards in the future.

In a study I did of the standards committee members and also of GHTF industry reps, the smallest industry rep had annual sales of 200 million dollars. Four members that authored 60601-1 3rd edition came from one company that had annual sales of 25 Billion and they were not even the biggest member. I threw out paid consultance, who are small business, but are normally hired by big companies. This profile is very typical of "industry" input, ...but I'm sure you already knew all that

Compare this to new industry statistics in Canada showing that 95% of medical device industry is small and only 1.5% are large, and you can see how a bulky and expensive standard, which is "small change" to a large company, suddenly becomes a barrier to trade to everyone else when thrust under a medical device regulation as "Mandatory". Fortunately, this standard isn't manadatory anywhere that I have seen (in a true regulatory context).

It is understood that the process of writing the standard is difficult. But as gramaley correctly pointed out, it still needs to be vetted, validated as being logically correct, effective at reducing risk, and with reasonable use of limited resources.

It's likely that $100M+ is being spent around the world upgrading to the 3rd edition. With that in mind, the excuse that "we tried hard" is not enough. Committee members are professionals, although they are "volunteers" they are sponsored by their company. It's seen as a privilege and you get kudos from being on the committee. So there is a serious responsibility here.

Also, we don't expect perfection. All we expect is being reasonable. There is plenty of stuff in the standard which is so far beyond reasonable that the criticism of "poor standard writing" is more than a sideline complaint. There is quite clearly a basic philosophical problem with the 601 series. I said before the problem is with the "IEC", I should clarify that this is just the 601 series.

For example, 601 clumsily introduces the concept of double faults in Clause 4.7, and the use of risk management to derive the faults. When a member of the 950 committee was asked about this, they laughed and clearly understood it made no sense.

The problem is, we need standards. If we didn't have standards, manufacturers would not even worry about being single fault safe (and many other things). But by introducing double faults, and by using vague references to risk management, we have ended up creating a situation where even good clean application of the simple single fault condition has got lost in a mess. As is so often the case, overkill in standards leads to underkill in practice. That is the big worry.

The problems with Clause 4.7 should have been obvious from the start. The people that wrote the clause should have had in mind how it was going to work, should have prepared practical examples, tested the clause as being effective. Deriving faults from "risk management" is not effective because 14971 does not have the tools to do this. Try it. 14971 just documents the results, not the methods. And double faults are so rare that it makes no sense to consider them in general; they should be treated as a special case only.

So, yes writing standards is difficult. But that's no excuse.

I have received a noncompliance report from our NB. The report said there is no evidence to show that the certified products comply with EN 60601-1:2006. Is it appropriate?