When to Validate and when not to Validate

[The Specialist]

Re: When to Validate and when not to

Thank you all for responding...
here's another layer... some pharma customers don't even audit for validation. At all. They scratch it out of their plan.
Others just care if we qualified the equipment.
And this is the root of the confusion here... when our customers can't even get on the same page.


Ronan - We do verify. We have an first piece inspection, in-process inspection, and final inspection plan that is fairly robust benchmarked against our competitors.

Gert - Seems to me the FDA is more concerned with the copy that goes on to the paper than the shape of the folded paper itself. The part never touches the drug. It either gets stuck on a bottle, or placed in a box. What you say is part of the problem for drug company suppliers like us - the drug companies expect us to do what they do... but the margins in a printed insert or outsert don't really allow for the techinical staff that pharma companies are accustomed to haveing on hand. If we try to increase the cost to allow for these resources we lose the business. It's a hard place.

Yodon - the specification is +- 1/16 all around. Not hard to verify. I know it's effective because we've never had a complaint about the length/width/thickness being out of spec going back through 4 years of records.
In fact the only complaint about this particular process is missing glue, and off counts. And even then it's under 4 PPM.

Illuminatus,

I work in the Pharma/BioPharm industry and I can tell you that the only things I would care about from a regulatory/validation point of view (for your paper insert) is:

The Print Quality
The accuracy of the variable data
The accuracy and readability of barcode/datamatrix/pharmacode

Beyond those requirements I would want to have a specification for size and weight of insert.

The size is important for three reason:

Fit in our containers
Meets design template (+/- agreed tolerance)
Control of weight (specification)

The weight is important only for the pharma company's packaging process. - The presence of an insert in the final pack is determined by final packaged weight measurement.

The ‘fold’ of the insert would be of no interest to me whatsoever. I would personally therefore not insist on validation of the process you describe. But, as others have said, if your customer ‘requires; it… what can you do?!

 

[Illuminatus]

Re: When to Validate and when not to

I would expect FDA to be more concerned with the paper itself, the contents of the paper, and the controls regarding that content. I would focus the validation on those issues.

Well - the paper is generally specified by the customer so we just show proper traceability and receiving processes. And remember - we're just talking about the folding part here...the printing is another process that we have validated (as Kevin pointed that we do have a validation SOP in place)

If you already have that information use that to support your validation, it all goes to show that the process is in control. Validation should be risk based, and the risk of the paper is not necessarily whether it is within the specification of size, it is more whether the contents is correct, legible and in control.

This is what I think I might need to formalize. The risk is minimal. In fact I recently did an FMEA as part of a quote package for a device company and the RPN numbers were all low.

 

[Illuminatus]

Re: When to Validate and when not to

Illuminatus,

I work in the Pharma/BioPharm industry and I can tell you that the only things I would care about from a regulatory/validation point of view (for your paper insert) is:

The Print Quality
The accuracy of the variable data
The accuracy and readability of barcode/datamatrix/pharmacode

Beyond those requirements I would want to have a specification for size and weight of insert.

The size is important for three reason:

Fit in our containers
Meets design template (+/- agreed tolerance)
Control of weight (specification)

The weight is important only for the pharma company's packaging process. - The presence of an insert in the final pack is determined by final packaged weight measurement.

The ‘fold’ of the insert would be of no interest to me whatsoever. I would personally therefore not insist on validation of the process you describe. But, as others have said, if your customer ‘requires; it… what can you do?!

I've not yet seen weight on a specification, so we don't weigh the pieces nor have we been asked.

Here's the real hard part.... and this may open a whole other discussion... I don't even know if the customer requires it.
None of the quality or supplier agreements or specifications say that the product must be made on validated equipment.
The purchasing people I've spoken to don't seem to care.
The incoming QC people don't care because it's not noted on a C of C.

It's the auditors that keep hammering this.

I honestly believe that many of our customers (and they are the biggest of the big pharma) have very poor communication between supplier audit-purchasing-QA.

 

[The Specialist]

Re: When to Validate and when not to

I've not yet seen weight on a specification, so we don't weigh the pieces nor have we been asked.

The size and thickness (specification) would determine the weight. We (the end user) would determine the weight for our own purposes...

I am not suggesting that you would weigh them!

Although, reading back over my previous post... that is what I implied, I realise!!

 

[medical_eng]

Re: When to Validate and when not to

I read this thread started by Illuminatus, plus I've read many other threads on validation here in the Cove (excellent resource).

So, the tendency that I'm noting is for everyone to default to (and it's good to take the high road, don't get me wrong), if there is the slightest of doubt then validate. But really, are we not tending to overcomplicate things a little by doing so?

The GHTF document on process validation clearly contemplates several situations where verification, if done well, is totally adequate (pg 7), and ISO9001/ISO13485/CFR820 all imply (to me at least) that the requirement for process validation is not mandatory across the board.

It would seem that Illuminatus has a perfect example of where validation is perhaps overkill. Save for the auditors, everyone else in Illuminatus's world (by his account) appears to share this view.

As a counterargument, why don't we just make the rule 'validate everything'? All the time, every time, no matter what, no exceptions. It'd be an easier rule to follow, eliminate confusion, and besides, (slight sarcasm here) can you ever be really, really, absolutely positively sure of your process output without 'validation'?

Anyone willing to offer clear examples of where process validation is specifically not done, that the decision not to validate is documented (and where such documentation should reside and perhaps the risk analysis/decision tree which substantiates this decision), and the verification done in lieu of validation, all compliant with the spirit and intent of the ISO/CFR regs?

Cheers!

 

[Illuminatus]

It's not that we're overcomplicating things...Validation for what we do really isn't that complex... it's that businesses like mine often don't have the resources to initiate and maintain a validation master plan for everything. We would have to pass that cost on to the customer.

If I went to a purchasing agent and said "your auditors are writing us nonconformances for not validating our folders so we are increasing your piece price to pay for a quality engineer"... well how do you think that would fly?

Now - if we land a new account and they say "in order to get the business you have to validate the equipment that our product is to be run on" at least we can build the man hours into the price of the piece.

Its the after the fact observations that frustrates more than anything.

 

[medical_eng]

Hi Illuminatus,

Sorry, I didn't mean to imply that you specifically were overcomplicating things. My post, admittedly changing the direction of the thread a little, was to call out the need to validate everything all the time as it seems like your auditors were defaulting to.

I read somewhere in a FDA document, correct me Covers if I am mistaken, that the FDA supports the 'least burdensome approach' while still requiring companies to meet all the requirements of CFR820. Not validating a process when it can be shown that verification of the output is sufficient would then be the correct approach, would it not? There seems to be very little guidance out there on how do to that well, with practical examples.

Cheers!

 

[Ronen E]

There seems to be very little guidance out there on how do to that well, with practical examples.

Hi,

I'm not sure what "that" you referred to. Typically lack of guidance isn't the problem with the FDA.

And to your main point - whether to validate or verify (when both are applicable) is many times a matter of cost. When running small production volumes of a complex device it usually can't pay for validation.

Cheers,
Ronen.

 

[medical_eng]

Hi Ronen,

"That" refers to my prior sentence: "Not validating a process when it can be shown that verification of the output is sufficient" Hopefully it clears it up for you.

Sorry for hijacking your thread Illuminatus but it would seem to me, from what you describe, that the only sane and logical task would be verfication only.

Cheers!

 

[jscholen]

I would say that your auditor is over stepping their authority. They should be auditing you against their quality requirements spelled out in a quality plan or supplier agreement.

If there are no requirements other than the specification, then you could compromise with her that you will correct your procedures to provide for justification when validation is not required since your process is clearly verifiable from a folding and packaging perspective and that you follow industry standards for this type of operation.

If she needs the process to be validated, then she needs to initiate a written agreement in conjunction with her purchasing and quality department on the requirements for validation because that could be a slippery slope of "how much" do you validate.