8 replies to this topic

[Annguyen]

Hi All, 

I have implemented a risk assessment plan that provides sampling frequency and number of samples depending on each production area. For example, high-risk areas take 3 samples/week, low-risk areas take 2 samples/2 weeks.

But the auditor asked why they chose 3 samples and not another number. Can someone help me answer the question or is there any standard for choosing the number of swabtest samples

Thanks a lot!

[MOHAMMED ZAMEERUDDIN]

You can justify that you have established different sampling frequencies and numbers of samples based on the risk level associated with each production area. High-risk areas, where there is a greater potential for contamination or quality issues, have been assigned a higher sampling frequency. This increased frequency allows to closely monitor and mitigate any potential risks in these areas.

On the other hand, low-risk areas, where the likelihood of contamination or quality issues is lower, have been assigned a lower sampling frequency. This frequency helps an adequate level of control and assurance over the quality of products, while optimizing resource allocation.

[Annguyen]

Thank you, I answered but the auditor asked to prove why 3 samples were chosen, and not less or more. While there is no standard that provides criteria for how much quantity to choose?

[Dorothy87]

Hi Annguyen :)

 

Not sure what type of risk assessment you have, but the simplest way is to split it for a non-food contact and food contact surfaces. Non-contact like floor, table legs etc are low risk, where table (food contact) machinery (food contact - blades) are high risk

. 

For example, you have 5 tables, they are all the same, then only one shall be swabbed. Slightly different will be when you have different types of machinery. Each shall be assessed for the worst case scenario,  if they are same, then just pick one as a representative. 

 

[vuongnt]

Hello An from Việt Nam,

 

In my opinion, i think that there is no regulation to define how many samples that you need to have do swabtest. All should be based on risk assessment and you can add more informations below when you answer auditor:

1. Add historical data to assess risk and you can compare the current year's swab testing plan with the previous year.

2. Number of sample is not fixed number. When contamination is detected --> risk changes --> increase the number of test samples and test frequency

3. The frequency and number of samples are naturally affected by the economic factors of the business. Therefore there must be a level of risk tolerance...

 

I hope this can help you a bit!

 

Good luck!

[Annguyen]

Yes, I understand. Thank for all

[kingstudruler1]

I wrote all of this and then realized Mohammed basically stated the same idea.   I think we are both saying the same thing.  

 

 

I think that is the problem with just declaring "X" number of samples for each zone.   Or, even using sites picked at random.   While these are sometimes accepted during audits, there is certainly room for interpretation as to how risk was used to determine them and usually leave a lot of room for risk.  

 

What I have done in the past is to create a register of all possible areas to sample . All zone 1, zone 2, zone 3, and zone 4.    You can further break it down by difficulty to clean as is shown in the 3m environmental monitoring handbook.  

https://multimedia.3...ng-handbook.pdf

 

The highest risk areas (zone 1 or zone 1 hard to clean) would be assigned the highest frequency - like weekly.   The second highest risk areas (zone 2 or zone 1 easy to clean) would be assigned a frequency less than that of higher risk areas - in this case call it monthly.    Then continue until all risk areas are assigned a frequency.   When areas test positive then can be moved to a higher risk category.  You would still have the opportunity for someone to questions why you chose "X" frequency and not something else.  IMO people are less likely to question this as opposed to random numbers of samples and /or sites. 

[Leila Burin]

Hello,

you could also answer based on stats.

If you want to demostrate a certain Cpk, then n=30 shall be the minimum samples to test

 

best regrds,

Leila

[Leila Burin]

Hello again,

I found this reference:

 

"ne approach to determination of the number of sites would be to address it in an manner similar to that of ISO 14644-1 for the walls and floors (as relevant). Each surface would then be treated as a separate item and the minimum number of sites determined for each as the square root of its surface area."

 

And is also commented in a Eurofins webinar (https://www.eurofins...tal-monitoring/), minute 35.49

 

best rgds,

Leila