How to set Cleanroom Alert and Action Limits as part of Microbiological Evaluation

[BioBurden]

Hi Guys,

Back again to pick your brains.

We have a class VII cleanroom, approx. 4.5m x 4.5m.

We perform environmental monitoring on a monthly basis, "at rest".

Air plates, settle plates and contact plates.

Our limits are as follows:

Air Plates in Cleanroom - 100 (cfu/mould per m^3)
Air Plates in Airlock - 200

Settle Plates in Cleanroom - 100 (cfu/mould per plate)
Settle Plates in Airlock - 200

Contact Plates in Cleanroom - 25 (cfu/mould per 24cm^2)
Contact Plates in Airlock - 50

I know we should be testing while in use which is what we plan to do shortly.

However, I have been asked to set alert limits (below the above limits) and I have been unable to find any helpful information regarding any formula or method to define these limits.

Can anybody help?

 

[Ajit Basrur]

Re: How to set Cleanroom Alert and Action Limits as part of Microbiological evaluatio

Did you refer an existing thread on similar subject -

Establishing alert and action limits for clean room bioburden

 

[BioBurden]

Re: How to set Cleanroom Alert and Action Limits as part of Microbiological evaluatio

So here are some of the calculations I have done based on the following historical data:
AIR PLATE SDA (mould/ m^3) - Class VII Cleanroom, at rest
JUN - 2
JUL A - 1
JUL B - 1 *2nd sampling performed due to spike in one area,
AUG - 22
SEP - 0
OCT - 0
NOV - 0

Mean: 3.714285714
STD DEV: 8.097618698
Pop STD DEV: 7.496938151

95% Confidence Interval (WARNING)
3.714285714 +/- 5.998689651
Upper Limit: 9.712975365 (Rounded up to 10)

99% Confidence Interval (ACTION)
3.714285714 +/- 7.883614549
Upper Limit: 11.59790026 (Rounded up to 12)

Does that seem right?

That would indicate that an investigation into August's results would have been necessary at the time despite it being well within the 100 specified limit for Class VII as set out in Eudralex.

Where am I going wrong?

 

[BioBurden]

Re: How to set Cleanroom Alert and Action Limits as part of Microbiological Evaluatio

Am I supposed to be adding 2 or 3 Standard Deviations instead of the CI?
Have I just confused the situation by adding the CI into the calculations?

 

[Ajit Basrur]

Re: How to set Cleanroom Alert and Action Limits as part of Microbiological Evaluatio

Hi Bioburden,

For Poisson-distributed data, the alert limit may be set at the 95th percentile of the dataset, and the action limit at the 99th percentile.

For normally distributed data, the limits may be set at the average plus 2x standard deviation (equivalent to 95% probability) and 3x standard deviation (equivalent to 99.7% probability) respectively.

Refer to this execllent document for more info - http://www.pqg.org/pharma/wp-conten...trol-for-Non-sterile-Pharmaceuticals-2008.pdf

 

[BioBurden]

Re: How to set Cleanroom Alert and Action Limits as part of Microbiological Evaluatio

So for the above example, I was correct in using the CI values to calculate?
It's definitely not normal distribution.

Sorry, just to clarify, I have no SPC software, I am just using Excel.

 

[Ajit Basrur]

Re: How to set Cleanroom Alert and Action Limits as part of Microbiological Evaluatio

So for the above example, I was correct in using the CI values to calculate?
It's definitely not normal distribution.

Sorry, just to clarify, I have no SPC software, I am just using Excel.

No issues ... you could take 2 SD and 3 SD for alert and action limits resp.

 

[BioBurden]

Re: How to set Cleanroom Alert and Action Limits as part of Microbiological Evaluatio

Hi, just a quick question, returning to the decision process for Poisson or Normal.

I'm struggling to identify which distribution is which when charting the results using excel.

From the above document:
"The data should be plotted graphically to ensure the correct distribution is chosen for modelling. If the distribution is unknown, then the use of cumulative probability will allow limits to be calculated"


Also, from the above document:

"The decision regarding the approach to setting the appropriate limits should be captured in a written justification document."

What would this document need to include?

 

[BioBurden]

Re: How to set Cleanroom Alert and Action Limits as part of Microbiological Evaluatio

Any answers for the previous post?

I have some more questions.

At what stage do I stop updating the limits based on historical data?

Once I have 12 months of data? i.e. 12 results.

Otherwise, if I keep using +2SD and +3SD, the process never goes out of control.

Does that sound right?