IEC 62304 Amendment 1 published
By Mitch on Friday 10 July 2015, 11:52 - Standards - Permalink
The new version of IEC 62304, also known as IEC 62304:2015 or amendment 1 of IEC 62304 was published by the IEC at the end of June 2015.
There were no major changes compared to the drafts that were circulated earlier this year.
The two major new requirements, compared to IEC 62304:2006 are:
- Requirements about legacy software,
- Changes in the definition of the security classes, based on risk assessment.
IEC 62304:2015 is available on IEC website at the astounding / amazing / appealing / astonishing (delete as appropriate) price of 650 swiss francs (approx. US$700) for the consolidated version.
Now we need to wait for this version to be harmonized by EU and recognized by the USA.
Thanks for the update.
Do we have an estimation of the application date of the new 62304 version ?
Also, do you know if we should face evolutions in the final version compared to this red-lined version?
Thanks for your blog,
There is no agenda published neither by the FDA nor by the EU for the application of new versions of standards. But we can reasonably estimate that IEC 62304 2015 will be harmonized in the EU and recognized by the FDA within six months to one year.
This version is the final version of IEC 62304 2015. The only possible changes are the ZA, ZB and ZC annexes that the EU could add on the harmonized versions of standards. But it won't change the technical content of the standard whatsoever.
Thanks for your feedback.
Is there a reason that IEC 62304:2015 has not been harmonized yet?
Should I be wary of using it on my Essential Requirements Checklist (IVD - General class)?
It's not harmonized yet, like about a few hundreds of new versions of other ISO and IEC standards. The reason is the turmoil generated by the new european regulation 2017/745/EC and 2017/746/EC. The list of harmonized standards haven't been updated since then.
You don't have to be wary of using new versions. Rather, the notified bodies "suggest" to manufacturers to use the lastest versions as they are state-if-the-art.
Thanks for all your articles!
Did the specification "If the HAZARD could arise from a failure of the SOFTWARE SYSTEM to behave as specified, the probability of such failure shall be assumed to be 100 percent. (section 4.3)" disapear with the revision?
Yes, it disapeared!
But considerations about the probability of software failure are still present in the informative section B.4.3
also keep in mind that 100% probability of software failure is often not equivalent to 100% probability of resultant harm.
logically, it makes sense as well to add an external (to software) risk mitigator if software failure alone has a risk of serious harm.