Future of ISO 14971 - Possible revision?

[Marcelo]

ISO TC 210 JWG 1 will meet next week in Tampa, Florida, to discuss the comments on the systematic review of ISO 14971. Although the voting by NCs indicated the desire to maintain the current standard as is, the number and types of comments were deemed enough to require a careful analysis before a decision is reached.

Also influencing this is the fact that ISO finally removed ISO 14971 from the list of MSS standards (ISO 14971 was erroneously put there), which means that ISO 14971 won't need to follow the HLS anymore. This is one of the issued that were impacting any revision (the other being alignment with ISO 31000).

 

[Marc]

Is the current revision 2012?

MSS = Manufacturers Standardization Society?

HLS = ???

 

[Marcelo]

Current revision is 2007, the 2012 version is the EN European one with the weird technical deviations.

MSS = Management System Standard

HLS = High Level Structure for MSSs (the structure of SO 9001:2015)

 

[Marcelo]

ISO TC 210 JWG 1 met in Tampa last week to discuss the comments from the systematic review of ISO 14971.

The meeting was really good, we did not discuss all the comments, but the ones that were perceived as the most critical.

In general, the consensus was that the standard text is is principle ok (although sometimes in a too general level) and most of the comments were related to possible guidance. We discussed how to include guidance, in the standard or in a revision of ISO TR 24971 (although it was pointed out that a lot of people do not know 24971 exists - common in the case of guidances, unfortunately).

One specific point that needs more input for a future decision is the revision of the MD directives in Europe (and we should probably have a revised text soon).

The group will meet again in October together with the IEC TC 62 plenary meeting. There a decision of a suggestion will be forwarded to the parents committees.

 

[Ajit Basrur]

That's true, Marcelo. I did not know about the ISO/TR 24971 for a long time

 

[Marcelo]

Two topics which Brazil (and other countries) commented were:

- The need to downplay the examples of risk matrices in the standard (a lot of people confuse the matrix, which is a risk ranking tool, with the criteria or even the policy for risk acceptability, and also people usually simply copy a matrix from the standard without understanding that that means nothing - in fact, most of the people I know have no idea of how to properly construct a risk matrix - from the risk plane to iso-risk contours to deriving the matrix from these).

- The need to create a process for legacy devices. Although ISO 14971 is somewhat old, a lot of manufacturers, in particular in countries with no or low tradition of risk management, have older (and even newer) devices which do not comply with ISO 14971, so now that their regulatory systems are evolving, the regulations are asking or expecting compliance with ISO 14971 (or risk management in general). Not having this even mentioned in the standard creates a problem for all those trying to apply the standard for legacy devices.

 

[Mark Meer]

....Not having this even mentioned in the standard creates a problem for all those trying to apply the standard for legacy devices...

Why is applying ISO 14971 retroactively to legacy devices a problem?

How would the framework be any different for a legacy device? Wouldn't all the same principles apply?

...most of the comments were related to possible guidance...

...so, in otherwords, the standard itself isn't likely to get revised any time soon...?

...The need to downplay the examples of risk matrices in the standard...

This can presumably just be addressed through detailed guidances, without necessitating a revision to the standard itself, no?

----
Curious: other than the above mentioned comments, was there anything else that came up? ...I'd be especially interested in anything that might, in the future, be considered/included in a future revision.

 

[Marcelo]

Why is applying ISO 14971 retroactively to legacy devices a problem?

How would the framework be any different for a legacy device? Wouldn't all the same principles apply?

Not exactly different, but it may be better to change some of the order of the steps (and also not use some). For example, a legacy device already has "risk control measures" included, so it would be easier to reverse engineering what is already there, and then check if there's anything new. You can check the attached document for an example of a possible different process.

...so, in otherwords, the standard itself isn't likely to get revised any time soon...?

It still may, there were some itens there were perceived as requiring a changing in the requirements of the standard. Also, we are still discussing if, even if we are left with only guidance, it should be in the standard or not.

This can presumably just be addressed through detailed guidances, without necessitating a revision to the standard itself, no?

Yes, but it really depends on further discussion. My opinion is that we should revise the risk acceptability criteria requirements to make them more clear, this would also help with the matrix problem.

Curious: other than the above mentioned comments, was there anything else that came up? ...I'd be especially interested in anything that might, in the future, be considered/included in a future revision.

There were 40 pages of comments, so yes, there were more stuff discussed.

 

[medical_eng]

Hi Marcelo,

Reviving a somewhat old thread, has the revision to ISO 14971 actually included guidance for legacy devices? There's info on the web as to the known changes, but the info is silent on this.

By the way, very helpful guidance in your attachment, and actually makes a little more sense than the guidance offered in 62366 for usability of unknown provenance.

Thanks!

 

[Marcelo]

The topic was transferred to ISO TR 24971 and revised, but is still included in the latest draft, in a revised form (without the flowchart and table, but in essence it's the same guidance).