ISO11135: Ethylene Oxide Sterilization - Number of BIs for routine monitoring

[123gkb]

Hello everyone

Info:
- We have a 15m^3 EtO sterilization chamber, with a usable volume of 12m^3.
- Our current process spec validation (and subsequent requalification reviews and reduced requalifications) has been audited with no findings for 5+ years.
- We currently use 18x BIs for routine monitoring, as per the recommendation of ISO 11135:2014 Table C.3.
- (According to the table, we can actually use 16, corresponding to the usable load volume)
- BIs are placed in validated IPCDs for routine monitoring
- In 2000+ cycles over 10 years, we have never had a postive BI or device testing non-sterile.
- We have a 3h exposure time. With fractional cycles we have all-kill of 36 distributed BIs and inoculated devices at 30min exposure.

My question:
- Annex C is marked as "informative", and not "normative"
- Table C.3. is titled "Recommended number of BIs", and not "required"
- It seems that the number is not an explicit requirement of the standard, provided that a reduction in the number is robustly justified
- Can we use historic data, data from fractional cycles, temperature/RH profiles etc to justify a reduction in BI's, and limiting them to the 3x most challenging locations?

Reason:
- We do 30+ cycles per month, and the cost of BIs per cycle is substantial
- Our devices are low cost, but large - thus the production cost is heavily influenced by cost per cycle.
- 18x seems excessive considering we have never had a failed BI using this process spec in this chamber.
- As mentioned, the standard "recommends" the number in an "informative" Annex

Thanks in advance!

 

[planB]

ISO 11135:2014 Table C.3 contains the gold standard for the number of BIs. Nobody will question you as long as you follow these numbers. So in case you chose to reduce the number of employed BIs, be prepared to be questioned and challenged frequently.

Yes, you may reduce the number of BIs, based on a strong rationale.

Regarding your (history of) sterilization runs and product:

1) How many sublethal and half cycles have you performed for your current cycles over time?
2) What is your frequency and extent of periodic re-validation?
3) Are you sterilizing in dedicated or mixed loads?
4) Is your load rather uniform or vary a lot from run to run?
5) How would you rate your product in terms of being a challenge to the process?
6) What regulatory device classes are your products belonging to?

 

[123gkb]

ISO 11135:2014 Table C.3 contains the gold standard for the number of BIs. Nobody will question you as long as you follow these numbers. So in case you chose to reduce the number of employed BIs, be prepared to be questioned and challenged frequently.

Yes, you may reduce the number of BIs, based on a strong rationale.

Regarding your (history of) sterilization runs and product:

1) How many sublethal and half cycles have you performed for your current cycles over time?
2) What is your frequency and extent of periodic re-validation?
3) Are you sterilizing in dedicated or mixed loads?
4) Is your load rather uniform or vary a lot from run to run?
5) How would you rate your product in terms of being a challenge to the process?
6) What regulatory device classes are your products belonging to?

Thank you for your detailed answer!

1.) We have done 2x sub-lethals and proved D-value for IPCD >> that of product. We do one half cycle at least every two years, but also do half-cycles with introduction of any new equivalent variants of a device family. I would say about 10 half cycles (don't have access to data right now). We have never had a positive IPCD for half-cycles or routine cycles.
2.) We do a yearly review of equipment, data trends, bioburden (sampled quarterly), any device changes etc.
We do a half cycle at least every two years (requirement of review protocol)
3.) Mixed loads
4.) Load varies in terms of types of devices / device families. However, the devices are very similar in construction, materials and load density. Validation was done with high density mixed load
5.) Very low challenge to sterilization. Large airways, no absorbent materials, rigid polymers. Our routine bioburden levels are also very low. For validation of IPCD we inoculated devices with BIs placed in most challenging locations in device. For the worst device, we got 9/10 kill for 20min fractional cycle. Routine exposure is 180min (600mg/L).
6.) Device classes: 3x Class 1s, and 1x Class 2a

 

[planB]

A history of 10 half cycles seems not really a lot to me. However, it seems to me, that your minimum half cycle length would be 30 min, allowing for a full cycle of 60 min. Since you have set your actual full cycle to 180, your log-reduction is 3x12 = 36, adding a considerable additional margin of safety.

Could you also save cost by reducing your EO dwell time by 2h, and thus, maybe even save aeration time?

In case you still want to reduce the number of BIs, I would suggest aiming at 8 BIs (i.e. half of the recommended number): you would still have to map the entire usable chamber volume, not just a worst-case location, because this is the intention of the in-process controls: demonstrating adequate lethality in the while usable volume. In return, would you be in a position to increase regulatory acceptance by increasing your re-qualification frequency, i.e. going to an annual performance of a sub-lethal or half cycle?

Another idea: Since you mentioned also monitoring of physical parameters: would measuring humidity and EO concentration in addition also be an option, allowing you to release product parametrically without any BI at least a week earlier compared to a BI release?