Key takeaways
- There is no universal limit: a threshold must be linked to the product, the process and the site history.
- An environmental result only makes sense if it is obtained in the right place, at the right time and with an appropriate method.
- Monitoring hygiene indicator microorganisms, detecting biofilms and verifying cleaning are complementary, but they do not serve the same purpose.
This article is also available as a podcast in more than 20 languages. Click the ⚙️ Settings icon on YouTube, then choose Audio Track and Subtitles to watch it in your preferred language.
For a pathogen, the objective seems relatively clear: it should not be detected where it could threaten the product.
For hygiene indicator microorganisms, the situation is much less straightforward.
How many CFU are acceptable in the air or on a surface? Should the same limit be applied in every facility? And what is the value of a threshold that triggers alarms every week but is never associated with a product defect?
Conversely, can we consider an environment to be under control if environmental results remain low while products regularly show microbiological defects?
We asked François Bourdichon how to establish limits that actually make sense.
What is the purpose of testing for hygiene indicator microorganisms if they do not necessarily represent a health risk?
François Bourdichon: Hygiene indicator microorganisms do not necessarily block product release and do not always have a direct impact on product safety.
They mainly provide information about the level of hygiene control throughout the production chain.
An increase therefore does not automatically mean that the product is dangerous. However, it may reveal a deviation in the process, cleaning or the production environment.
That is why I do not look only at an isolated result. I look at how it changes over time and how it relates to what we observe on the finished product.
Are there regulatory limits for hygiene indicator microorganisms in the production environment?
No, there is no single value that would be suitable everywhere.
I always compare the level observed in the environment with what I find on the finished product and with what is acceptable for that product.
The same result will therefore not necessarily be interpreted in the same way depending on the product being manufactured, the process used, the sampling time and the location where it was obtained.
The threshold should help detect a real deviation. It should not simply generate alarms with no consequence for the product.
How do you approach air monitoring?
When we monitor the air, we use an active air sampler whenever possible because it provides a quantified result.
We also still use settle plates extensively because they remain practical.
But the number alone is not enough. Sampling needs to be performed at a time when the air can actually contaminate the product.
For example, during soft cheese production, if I find certain yeasts or molds in the air and then find the same microorganisms on the finished product, I can establish a link and then take corrective action.
Conversely, if nothing is detected on the plates while the product shows defects, I am probably not looking in the right place or at the right time.
And if the plates are heavily contaminated without any impact on the product, the selected sampling point may not be relevant.
We should therefore always look for a correlation between what we observe in the environment and what is actually happening with the product.
How do you establish an initial threshold in a new facility?
I start by setting an initial level, then adjust it based on experience and site data.
If the threshold is so strict that the alarm is constantly triggered even though the product remains under control, it may be too low.
Conversely, if there is no warning signal in the environment while the product eventually starts going wrong microbiologically, then the threshold, sampling point or sampling time are probably not appropriate.
For some counts, such as Enterobacteriaceae, we may for example start with a level of around 100 CFU and then adjust it.
But we need to be very careful with this example: the unit depends on the type of sampling and the method used. This is not a universal limit that should automatically be applied to every plate or every surface.
The threshold should then evolve based on the correlation observed between the environment and the finished product.
Are microbiological counts enough to assess the hygiene of a surface?
Not always.
For surfaces, we can collect samples using swabs, wipes or other sampling tools. But we are also increasingly looking for the possible presence of biofilms or residual contamination.
There are solutions based on UV lamps or colorimetric methods, for example.
At Sodiaal, we have purchased UV lamps for some sites. They can be useful for examining areas that we thought were dry, less exposed or did not necessarily require frequent cleaning.
This can sometimes make residues visible that we would not otherwise have detected.
These tools do not replace microbiological testing. Instead, they help draw our attention to certain areas and give us a better understanding of the real condition of the production environment.
Should environmental monitoring be distinguished from cleaning verification?
Yes, they have two different objectives.
When we verify cleaning, we want to know whether a surface has been properly cleaned after the cleaning operation. ATP testing can, for example, be used to verify the effectiveness of wet cleaning.
But strictly speaking, that is not production environment monitoring.
Environmental monitoring is more about understanding what may be present on working surfaces and in the environment while the site is operating.
The question of neutralizers also depends on this objective.
If we take a sample after a disinfectant has been used, we need to make sure that residues will not prevent the target microorganisms from growing. We therefore need to use neutralizers appropriate for the disinfectant used.
But this does not mean that neutralizers should automatically be used for all environmental samples. It depends on the disinfectant, its residual activity, the surface and the method used.
It is therefore essential to start by defining what we want to measure: cleaning effectiveness or the microbiological status of the production environment.
Thank you to François for reminding us that a microbiological threshold only has value if it helps us understand and anticipate what is really happening with the product.
If you have any questions, feel free to leave them in the comments… we’ll do our best to answer them.
References and further reading
• Regulation (EC) No 2073/2005 — distinction between food safety criteria and process hygiene criteria
• ISO 18593:2018 — Horizontal methods for surface sampling
• ACTIA — Good practice guide for surface sampling
• EURL Listeria — sampling guidance based on a case-by-case and risk-based approach