Clean room revalidation frequency

[zeihr]

Hello,

I work in a medical device company with a clean room that classified as a Class 8 ISO. We conduct the revalidation annually, when all the production activities are stopped, which is right on the religious holiday. My boss want us to have the full holiday at least once a year, and started to considering to "reduce" the frequency of the revalidation, since it will be hard if we ask the management to totally stop the production activities beside that religious holiday. We have some plan that we only do the at-rest monitoring (not checking all the sampling points as we usually do in validation, only some points that used in monthly monitoring), then the validation will be held once in 2-3 years, or do the revalidation every 2-3 years by third party.

I was ordered to find any references that could support this plan, it will be good if it based on ISO standard, especially ISO 14644.
In ISO 14644-2:2015 point 5, it stated "Periodic classification test in shall be undertaken annually in accordance with ISO 14644-1. The frequency can be extended based on risk assessment, the extent of monitoring system, and data that are consistently in compliance with acceptance limits or levels defined in the monitoring plan."

Could anyone please explain what that statement means? Does that mean we can do the reclassification/revalidation not in annual basis?
*additional info, our products is a terminally-sterilized class II medical device.

 

[Nichole F]

That sentence in 14644-1 does mean you can adjust the frequency based upon internal justification. That justification should be based upon the stability of the environment, the risk of the area being out of specification, and the means employed to monitor the area between validations. A justification of wanting to decrease the frequency to avoid a production shut down will not be acceptable to any regulatory agency and does not conform with ISO 14644.

Keep in mind, if you do have an out of specification reading during your routine monitoring, you must assess the impact on all product produced since the last in-specification reading. Depending on your production levels, monthly may present a large risk for you. You can do in-process monitoring but this is can present a challenge to establish a base-line and set justified limits since your cleanroom will almost certainly not meet the ISO class 8 requirements in-use.

If you have sufficient historical data to show your cleanroom maintains a steady-state of in specification readings during routine monitoring and at re-validation, you will need to justify why decreasing the validation cycle does not preset a risk to the products being produced and that you have sufficient controls in place to detect any excursions. Ensure your test methods align with ISO 14644-2 and 14644-2.2.

 

[planB]

In addition, I would recommend to not open this box of worms of stretching the re-qualification period from annually to something longer. Have a close look at the section you cited:

The first sentence is crystal clear: You must do requalification annually. What follows is an ambiguous second sentence that would probably be considered bad standards writing nowadays:
- "extending" would more relate to a time period; you would either "increase" or "decrease" a frequency; but what means "extending" a frequency?
- "can" means in ISO language "being capable of"; what is actually meant is "giving you permission" and this is nowadays reflected in ISO standards by using the term "may"

So an authority will probably hold you accountable for the very first sentence in case you claim compliance to ISO 14644-2. Making use of the second sentence could turn into an uphill battle that quickly outweighs any benefits gained from extending the requalification interval.

HTH,

 

[zeihr]

That sentence in 14644-1 does mean you can adjust the frequency based upon internal justification. That justification should be based upon the stability of the environment, the risk of the area being out of specification, and the means employed to monitor the area between validations. A justification of wanting to decrease the frequency to avoid a production shut down will not be acceptable to any regulatory agency and does not conform with ISO 14644.

Keep in mind, if you do have an out of specification reading during your routine monitoring, you must assess the impact on all product produced since the last in-specification reading. Depending on your production levels, monthly may present a large risk for you. You can do in-process monitoring but this is can present a challenge to establish a base-line and set justified limits since your cleanroom will almost certainly not meet the ISO class 8 requirements in-use.

If you have sufficient historical data to show your cleanroom maintains a steady-state of in specification readings during routine monitoring and at re-validation, you will need to justify why decreasing the validation cycle does not preset a risk to the products being produced and that you have sufficient controls in place to detect any excursions. Ensure your test methods align with ISO 14644-2 and 14644-2.2.

We're planning to do a review based on historical data, precisely from the last 10 years validation data. We will use that as a justification to the top management. But there's a problem; the in operational validation and monthly monitoring results always having the sampling points that exceed the specification. I'm not sure if we can convince the top management by showing the in-operational results, although the at-rest results never exceed the specification. Is it still valid if we only show the at-rest data?

My boss told me that it doesn't care if our cleanroom are "dirty" while in operational condition, because we're a terminally-sterilized device, we can increase the sterilization dose. Idk but tbh I'm confused, is it okay if we're relying on sterilization dose? I think its going to affecting bioburden and endotoxin result, right? What will happened if we keep increasing the sterilization dose?

 

[zeihr]

In addition, I would recommend to not open this box of worms of stretching the re-qualification period from annually to something longer. Have a close look at the section you cited:
View attachment 31102

The first sentence is crystal clear: You must do requalification annually. What follows is an ambiguous second sentence that would probably be considered bad standards writing nowadays:
- "extending" would more relate to a time period; you would either "increase" or "decrease" a frequency; but what means "extending" a frequency?
- "can" means in ISO language "being capable of"; what is actually meant is "giving you permission" and this is nowadays reflected in ISO standards by using the term "may"

So an authority will probably hold you accountable for the very first sentence in case you claim compliance to ISO 14644-2. Making use of the second sentence could turn into an uphill battle that quickly outweighs any benefits gained from extending the requalification interval.

HTH,

I see. So, although we decided to decrease the frequency, the authority would still ask us why can't we do the requalification annually, right? They still expect us to do the "first sentence" first as long as we can do that.
If we stay insist on decreasing the frequency, should we do the risk assessment to justify that it won't significantly affect the product sterilization? AFAIK, we haven't received any complaints concerning product sterility until now.

 

[planB]

You can always opt to not follow a standard clause by clause, and a strong rationale built upon solid risk management could be the basis for this.

However, lack of complaints related to sterility failures is too weak of a basis to demonstrate that you have controls in place to manufacture product of adequately low bioburden, which is the input to your sterilisation process; lack of complaints as sole justification is more like this quote: "absence of evidence is not evidence of absence".

 

[zeihr]

You can always opt to not follow a standard clause by clause, and a strong rationale built upon solid risk management could be the basis for this.

However, lack of complaints related to sterility failures is too weak of a basis to demonstrate that you have controls in place to manufacture product of adequately low bioburden, which is the input to your sterilisation process; lack of complaints as sole justification is more like this quote: "absence of evidence is not evidence of absence".

I'm going to collect some records to support our proposal, such as validation report, bioburden test results, and monthly monitoring results. Is it enough? Or is there any other data I should prepare?

 

[chris1price]

I think its going to affecting bioburden and endotoxin result, right? What will happened if we keep increasing the sterilization dose?

There are a couple of points here, increasing the bioburden will mean you should revalidate the sterilisation process. Increasing the sterilisation dose may cause increased degradation of any plastics; you may have to repeat design verification testing. All additional time and costs to weigh against the change in cleanroom revalidation frequency.

 

[Nichole F]

We're planning to do a review based on historical data, precisely from the last 10 years validation data. We will use that as a justification to the top management. But there's a problem; the in operational validation and monthly monitoring results always having the sampling points that exceed the specification. I'm not sure if we can convince the top management by showing the in-operational results, although the at-rest results never exceed the specification. Is it still valid if we only show the at-rest data?

My boss told me that it doesn't care if our cleanroom are "dirty" while in operational condition, because we're a terminally-sterilized device, we can increase the sterilization dose. Idk but tbh I'm confused, is it okay if we're relying on sterilization dose? I think its going to affecting bioburden and endotoxin result, right? What will happened if we keep increasing the sterilization dose?

Your in-use readings will always be higher than at-rest. Most company's only set at-rest limits. Some may set operational limits, but those are tricky since your "Class 8" cleanroom is Class 8 at-rest more than likely. You are validating an at-rest state so using historical at-rest data is what you should be doing.

As for the second statement, that is entering dangerous waters. If the cleanliness of your cleanroom doesn't matter to your device, why use one at all? The company, at some point, determined cleanliness of the product was important, hence the infrastructure requirements. If your cleanroom isn't meeting the needs of the device, and you are relying on terminal sterilization to render the device within design specifications, you are taking a huge risk. Sterility isn't something that can be verified by your end user unless they are completing sterility testing when they receive the product. I don't know what device your company is producing, but unless there is some kind of bacterial indicator (I worked in culture media in the past; bacterial or fungal contamination was painfully obvious after 7-14 days), you won't receive complaints about a lack of sterility because it isn't detectable.

Your terminal sterilization validation was based upon a certain bioburden. Exceeding that bioburden means your sterilization process may not be effective in rendering the device sterile. You can absolutely increase the acceptable pre-sterilization bioburden, but you must validate it. When you increase the sterilization dose, you increase endotoxin risk. There is a balance that must be maintained. Also consider the risk if the device were to be released to the field without the appropriate SAL. As chris1price pointed out, increasing the sterilization dose may have an adverse impact on the materials used in your device meaning you may need to make design changes.

Make sure you have all of the information and have weighed the benefits against the risks before you attempt to decrease your validation frequency. The ISO standard prescribes annually for a reason. If you wish to deviate from that, you better have darn good evidence for it.

 

[zeihr]

Thanks for all the responses, guys. I'm bringing this whole thread while discussing with my boss, and it really helps us. I'll come here later if I have any other questions regarding this matter.