Verification that Manufacturing Equipment has maintained a Validated State

[dedvalson]

I am a QA manager with a growing medical device company, I have been tasked with setting up a workflow to verify production line equipment has remained in a validated state after it has been taken off the line for a repair.
Is there any guidance documents or industry standard practices which I could model my workflow from?

 

[Stijloor]

A Quick Bump!

Can someone help??

Thank you very much!!

 

[Ronen E]

Hello and welcome to the Cove

I'm not aware of any standard practice to achieve the above, and honestly, I don't think you need one. "Verify" is just a fancy word for "make sure", IMHO, and "workflow" just conveys an expectation (by whoever tasked you with it) that you explain how you're going to do it. In simple words: "show me how you're going to make sure that that equipment is still in it's validated state."

Of course, the simplest way would be to re-validate it. I'm quite sure you want to avoid that.

The other thing you could do is just "walk through" the last validation report, and make sure that all the stated and implied premises still hold, and that all the validation's outputs (parameter settings, control plans, other recommendations etc.) are implemented, as before the equipment was taken off for repair. Perhaps you should also prepare a monitoring plan (not an extensive one) for the first 1-3 months of operation post-repair, to avoid any surprises - weigh the risks and decide whether this is worth the resources.

Since a "workflow" was required, I'd just put all the above, in sufficient detail, in a written plan, and submit.

Please let me know if I completely missed the point...

Cheers,
Ronen.

 

[yodon]

Agree with Ronen (as usual). It's typical to do some kind of demonstration to prove the repairs didn't break anything. If it's too costly, the effort can be done (as Ronen suggested) with increased monitoring. The walkthrough of the last validation report is a good way to structure a risk-based decision on whatever you do end up doing. An FMEA assessing what could go wrong after / as a result of the repair is another good method.

 

[Mark Meer]

Depending on the type of equipment, it may be appropriate to keep samples of "known" quantities, and simply take measurements on these to ensure you get the expected results.

 

[dedvalson]

Thank you all for the suggestions, all the ideas given will be a big help to start implementing a workflow.

 

[Vthouta]

You could simply make a template for Verification and Validation form in a word doc :


Scope of Validation:
(Description of Change – Eg: A new equipment being used for first time, created new tooling for customer and validating the requirements, in-house fixture, etc)

Fixture Used for the validation (if applicable):
(Mention the asset number)




Verified the calibration of Fixture Used: (Yes/No)
(Document the cal due date of the fixture used along with Asset number)

Tools used for Validation:


Verified the calibration of tools Used: (Yes/No)
(Document the cal due date of the tools used along with Asset number)





Performed By: Date:
Validation Method:
(Mention the validation method used while performing the verification and validation)


Statistical sample size used for verification and validation:
Verification & Validation results:
(List down all the results of the verification performed that meets specifications and document the validation results by processing an ECN if required to prove the change)

Comments:





Dispositioned By:
Position: Date:

You can revise the format as per your needs and also could add the monitoring,etc.

Vik

 

[Vthouta]

Since you are working in growing medical device industry this could be a good read based on your application

http://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ucm064971.htm

For process validation - http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM070336.pdf

Based on this document, you can tweak according to your needs. I do completely agree with Ronen that there is no specific standard practice, but if you comply to FDA CGMP it would be a good practice to begin now and helpful in long run.

 

[Ronen E]

Since you are working in growing medical device industry this could be a good read based on your application

http://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ucm064971.htm

For process validation - http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM070336.pdf

Based on this document, you can tweak according to your needs. I do completely agree with Ronen that there is no specific standard practice, but if you comply to FDA CGMP it would be a good practice to begin now and helpful in long run.

I didn't refer to a "standard practice" for equipment validation (in fact I do know a few of those); rather, I was addressing what the OP was after - a "standard practice" for verifying that the equipment is still in a validated state (i.e. doesn't require re-validation).

Cheers,
Ronen.

 

[Grimaskr]

I'm not sure if it's applicable to the medical industry, but in various electronics manufacturing plants I've worked in we do a checklist and 'release to production' form after maintenance or engineering experiments have tampered with manufacturing equipment.

It's just a tickler list that asks questions like:

• Have all safety features (shield, guards, alarms, etc) been reactivated?
• Have all facilities (water, power, air, etc) been re-connected and checked?
• Have all process controls (vision systems, guides, etc) been checked?
• Does any of this maintenance/engineering activity require re-validation of product before release to production?
• And various other types of reminder questions.

The engineer responsible for the equipment needs to sign off on the checklist before Production starts running parts again. It serves as a good reminder system and shows auditors that you are double-checking things after tampering with equipment.

But again, I'm not familiar with medical industry rules/regs... so I don't know if this is applicable.