Package integrity Testing failure

[Balachander]

Hi,

Package integrity testing - Dye Penetration method for a medical device failed for real time aging tests. 3 out of 59 was a failure but the accelerated 1 year age testing (last year) did not have any issues.

1. My question is if based on the fact that accelerating aging tests for 1 year shelf life were passed a year ago. Can we perform Sterility test for these products (different batch) to check if the sterile barrier of the cartridge was or wasn’t broken , despite the fact of dye penetration failures in three pouches ?? Our investigation suggested we need to change the package design. We are doing this in order to launch the product on time and in parallel to implement design changes (include all required tests that should be completed).

2. My second question is if we redesign the package, this would affect the current sterilization validation right and it requires re-validation. Please let me know

Thanks,
Bala

 

[Ronen E]

Dye Penetration method for a medical device failed for real time aging tests. 3 out of 59 was a failure

What published standard are you applying? What is your statistical analysis method and pass/fail criterion? What is the statistical significance of this result?

the accelerated 1 year age testing (last year) did not have any issues.

Accelerated ageing is many times inappropriate. Assuming that you are using the simplified Q10 model, please note that it only applies to chemical reactions of order 0 or 1. I highly doubt that the deterioration of your barrier falls within this category (if it's even a deterioration - what were the results at 0, 3, 6 and 9 months?).

Can we perform Sterility test for these products (different batch) to check if the sterile barrier of the cartridge was or wasn’t broken , despite the fact of dye penetration failures in three pouches ??

You can, but (a) I highly doubt sterility could be maintained if the dye penetrated; (b) you'll need do test a batch that was packaged at the same time as the one that failed the dye test, and real-time aged under the same conditions; and (c) you'll have to apply very high statistical rigor (e.g. a big sample size) if you want to use a positive sterility outcome to overrule the negative dye penetration result - as a minimum.

if we redesign the package, this would affect the current sterilization validation right and it requires re-validation.

Correct.

 

[Tidge]

These strike me as the most important ("red-face") questions:

What published standard are you applying? What is your statistical analysis method and pass/fail criterion? What is the statistical significance of this result?

Even if 3 failures out of 59 samples was acceptable (in what I assume to be an attribute type test) for sterility (which I would assume to be implicated in rather high severity risks) strike me as being relatively low confidence with marginal tolerance (well below 90%). As far as risk profile for the device goes, we might as well not bother sealing the pouches since it wouldn't be appropriate to claim much of a reduction in risk (based on these results).

 

[Balachander]

What published standard are you applying? What is your statistical analysis method and pass/fail criterion? What is the statistical significance of this result?

Accelerated ageing is many times inappropriate. Assuming that you are using the simplified Q10 model, please note that it only applies to chemical reactions of order 0 or 1. I highly doubt that the deterioration of your barrier falls within this category (if it's even a deterioration - what were the results at 0, 3, 6 and 9 months?).

You can, but (a) I highly doubt sterility could be maintained if the dye penetrated; (b) you'll need do test a batch that was packaged at the same time as the one that failed the dye test, and real-time aged under the same conditions; and (c) you'll have to apply very high statistical rigor (e.g. a big sample size) if you want to use a positive sterility outcome to overrule the negative dye penetration result - as a minimum.

Correct.

Thank you Ronen for your recommendations.

two more questions:
1. We might not have products from the same batch that failed the dye test. Do you think using a different batch and testing it would be justifiable ?
2. With respect to sterilization revalidation: We would have to repeat the bioburden validation and find the new Correction factor right ?

Thanks in Advance

 

[Ronen E]

1. We might not have products from the same batch that failed the dye test. Do you think using a different batch and testing it would be justifiable ?

I didn't say it would have to be from the same batch. I said it would need to be (as a minimum, IMO) a batch packaged at the same time as the one that failed the dye test, and real-time aged under the same conditions.

2. With respect to sterilization revalidation: We would have to repeat the bioburden validation and find the new Correction factor right ?

Bioburden is more a function of the device manufacturing process than of the sterile packaging design. Do you intend to change the device design or its manufacturing process (other than the sterile packaging)? Anyway, you will need to follow the process in the standard.

 

[Balachander]

I didn't say it would have to be from the same batch. I said it would need to be (as a minimum, IMO) a batch packaged at the same time as the one that failed the dye test, and real-time aged under the same conditions.

Bioburden is more a function of the device manufacturing process than of the sterile packaging design. Do you intend to change the device design or its manufacturing process (other than the sterile packaging)? Anyway, you will need to follow the process in the standard.

We don’t have a batch packaged at the same time. This is a batch with a different timeline. Is this fine ??

Regarding the bioburden question- We intend to change the package design. Does this warrant to do re-validation wrt bioburden correction factor ?
Thanks,
Bala

 

[Ronen E]

This is a batch with a different timeline. Is this fine ??

By the sound of it, no.
Perhaps it can be justified somehow, but I can't say without putting some time into studying the details. Send me a PM for a confidential discussion.

 

[katesquire]

Hi Balachander - could you tell me what type of packaging you use that failed? I recently had 3 out of 30 packages fail the dye penetration test b (detected pinholes) and I found it extremely unusual. Curious if you had a similar issue. Ours also passed the accelerated aging tests.