EU Validation Requirement vs. FDA Validation Requirement - EN 62304

H

Hrobot

Dear all,

Our software development SOP (for a class 2 medical device with an embedded software) is based on EN 62304. I am now trying to submit a 510(k) and realised that the terminology for software testing requirement has changed from Unit Testing and verification/System testing to IQ/OQ/PQ (Installation/Operational/Performance Qualification).
Do we have to change our terminology documentation format to fulfill FDA's requirement?

The same goes to the final device functionality test protocol. Whilst this is fine for MDD do I need to create an IQ/OQ/PQ protocol for the US market?

Your advice will be greatly appreciated

H
 
Last edited by a moderator:

harry

Trusted Information Resource
Dear all,

Our software development SOP (for a class 2 medical device with an embedded software) is based on EN 62304. I am now trying to submit a 510(k) and realised that the terminology for software testing requirement has changed from Unit Testing and verification/System testing to IQ/OQ/PQ (Installation/Operational/Performance Qualification). ........

Somebody correct me if I am wrong. Unit testing applies to software - unit testing is a procedure used to validate that individual units of source code are working properly whereas IQ/OQ/PQ refers to the need to validate the complete equipment or machine at the relevant stages suggested by the term itself.
 
H

Hrobot

I think I made a mistake there to reference unit testing where I should have just referred to the final system testing.

I have checked the standards which are recognised by FDA and it includes 62304 Software life cycle, 60601 medical electronic equipment and 14971 risk management which are all the standards that we adhere to.

DOes this mean that I could use the Abrreviated 510(k) with a declaration of conformity to these standards?

regards,
H
 
W

Watchwait

I just went back through and read both the FDA 510(k) new Paradigm and the requirements for using an Abbreviated 510(k). Sounds like your case is a good example of why they created this abbreviated path! Good Luck!
 
W

Watchwait

I am also very interested in the interpretation of the IQ/OQ/PQ comment in HROBOTs first posting. Is the implication that if these particular terms are not mentioned in IEC 62304 that they are not required by FDA? I'm far from a software expert, hence my naivete/curiosity.
 

yodon

Leader
Super Moderator
I am also very interested in the interpretation of the IQ/OQ/PQ comment in HROBOTs first posting. Is the implication that if these particular terms are not mentioned in IEC 62304 that they are not required by FDA? I'm far from a software expert, hence my naivete/curiosity.

There is nothing in 21 CFR 820 (regulation by which FDA requires compliance for device development and manufacturing) that requires a particular test paradigm for product verification and validation.

In my experience, IQ/OQ/PQ are typically (only) used in manufacturing equipment validation, not product V&V.
 
A

Asherlee

I recently submitted a 510(k) for a medical device with embedded SW. I based my SW documentation and terminology on the outlines described in the FDA Guidance document entitled:" Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices, dated May 11, 2005".

From my experience with the FDA, when they publish a Guidance Document that is applicable to your situation, then you should carry out the recommendations contained in the Guidance Document as much as possible/applicable.

Note that if you submit Declarations of Conformity in lieu of actual test reports, then they should comply with the FDA requirements listed in the attached table.

Also, the FDA has several Guidance Documents on the Use of Standards in Substantial Equivalence Determinations.

Regards,

Asherlee
 

Attachments

  • Requirements_DOC.pdf
    8.4 KB · Views: 662
H

Hrobot

Hi all,
I just want to update on my findings after speaking to several regulatory consultants. That no matter if the standard is a recognised standard by FDA it is best to reword any test protocols, or reports to fit the guidance documents' terminology during the submission. Basically the easier we make for them to review the documents the better chances we have to get an approval.

Hrobot
 
W

Watchwait

Probably not the right specific thread, but I'll give it a whirl anyway. I came across a posting on another (related) site that I want to bring to our esteemed group of colleagues at The Cove:

Situation: The control software of an instrument is upgraded to a new version, and the vendor documents (probably release note, or software change history log) indicate that there are no significant changes (or
add some enhancements that we don't care about). Based on this information, it is appropriate to conclude that "the change has no impact on the functionality of the application and no revalidation is necessary based on the documentation provided."?

We have exactly the same situation and always agonize over how to handle them. On one hand (the conservative one) FDA says all changes to an approved device must be validated. Period. On the other hand, the "rationalization" that the vendor provided adequate documentation (and we audit the vendor and know his quality system is OK) would seem to be adequate justification for the "no validation required" position.

Thoughts? :evidence:
 
Top Bottom