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.
In a previous article, we explored with Pauline Hars how routine environmental monitoring is carried out.
In this article, we take one step back to understand how an Environmental Monitoring Program is actually built.
In other words: how do you define the overall EM strategy for a cleanroom? How do you qualify the room? How do you choose your sampling locations?
To answer these questions, we spoke with Cédric Astorg, who shares his expertise and walks us through, step by step, how to design a robust EM plan that truly fits real-world operations.
A dense interview… but a fascinating one.
Cédric, to start, could you introduce yourself in a few words?
Of course! Today, I’m responsible for sterility assurance. I’ve been working in the pharmaceutical industry for about ten years, with experience in biotechnology, solid forms, dermo-cosmetics, and sterile fill-finish.
One particular aspect of my career is that I’ve had the opportunity to help build two brand-new pharmaceutical sites. Everything had to be created, understood, tested… It was an incredibly enriching experience.
Cleanroom Qualification Steps
When a pharmaceutical company receives a brand-new cleanroom, can it start producing right away?
Well, no. You can’t start manufacturing batches for patients as soon as the cleanroom is delivered.
It’s a bit like getting a new car: before hitting the road, you check the inspection report, make sure everything works, adjust your seat, set up the dashboard, check that no warning lights are on, and make sure the driver actually has a license.
In the pharmaceutical industry, we must demonstrate that the cleanroom truly meets GMP requirements, especially Annex 1, which defines how controlled environments should be classified, and Annex 15, which describes the qualification process.
In short, you have to go through all qualification stages, from design qualification all the way to environmental performance qualification (EMPQ). The idea is to test, measure, and document everything before giving approval.
And only after completing all of this can you officially start producing batches intended for patients.
Can you walk us through the main steps involved in qualifying a cleanroom?
It all starts with one question: “What do I need?”. This need must be described in a user requirement specification (URS), which becomes the foundation for all qualification steps.
The first stage is Design Qualification (DQ). The goal here is to ensure that the cleanroom design, and all associated utilities (especially the Air Handling Units), complies with:
- what is described in the URS,
- what is required by GMP,
- and the site’s overall contamination control strategy.
Then comes Installation Qualification (IQ). This step verifies that all system components are correctly installed. This includes:
- checking the AHUs, terminal filters, and measuring instruments for airflow, temperature, humidity, and pressure,
- reviewing the furniture (for example, benches in the airlocks),
- and especially inspecting cleanroom surfaces, which must be smooth and easy to decontaminate. It’s essential to identify any potential areas where contaminants could accumulate.
In parallel, you can perform the Operational Qualification (OQ). At this stage, you must show that the cleanroom meets the required specifications at rest, meaning without any production activity.
As for the tests to be performed, nothing surprising here: at a minimum, you need to comply with the expectations listed in paragraph 4.25 of GMP Annex 1, which essentially provides a checklist of required tests.
So, when exactly does the EMPQ (Environmental Monitoring Performance Qualification) take place during cleanroom qualification?
Alright, so we have design qualification, installation qualification, and operational qualification… and at this stage, we still haven’t done any microbiological monitoring. This is where the EMPQ comes in.
During the EMPQ, we perform sampling for both non-viable and viable particles, and we also monitor environmental parameters such as temperature, relative humidity (RH%), and differential pressure between rooms.
The EMPQ demonstrates that the cleanroom can maintain the environmental quality defined in the user requirement specification, under real operating conditions. This means that:
- the equipment is already installed in the area,
- the flow management procedures (for people, materials, and equipment) are in place,
- the gowning procedure is in place,
- the cleaning and disinfection procedure is in place.
The EMPQ is carried out in two phases: static and dynamic.
Dynamic: performed under operational conditions, with simulated activities following the approved procedures, the maximum number of personnel in each room, and all production equipment running.
Static: performed at rest, meaning no personnel present and no production equipment in operation.
Once the EMPQ is completed, who decides whether production can begin?
At the end of the EMPQ, the qualification report must undergo a technical and quality review, and all identified deviations must be closed. After that, the cleanroom system can be formally released.
It is the Quality Assurance function, typically the Qualified Person (QP), Head of Quality, or Responsible Pharmacist who approves the use of the cleanrooms for GMP manufacturing.
“And contrary to a common belief we often hear in companies “Sedimentation plates outside Grade A are pointless” well, that’s actually false.”
Selection of sampling points and frequencies in the EMPQ
How do you choose the sampling locations to monitor?
The sampling plan must be based on a risk assessment. This is what allows you to identify the critical areas within a cleanroom. To choose the method, you can use FMEA or develop your own methodology.
Personally, I really like risk assessments structured around a series of questions that you answer with a simple yes or no.
Concretely, how do you do it? Can you give us an example of a method or thought process?
In this analysis, several selection criteria come into play:
Regulatory requirements, which depend on the nature and criticality of the operations performed in the area. Typically, you want to monitor locations where critical operations take place, such as open-phase filling or tubing connections.
The design of the room, taking into account the position of supply and return air vents. You need to ask yourself: “Is it more relevant to place a sampling point under a supply vent, or closer to the return, where contaminants are drawn in?”
Personnel and material flows. For example, in a large 300 m² room, you can create a spaghetti diagram to identify high-traffic zones and flow intersections, which are often more likely to show airborne contamination.
Gowning areas, which require special attention. First, because a lot of particles are generated there; and second, because you must ensure that the new layer of gowning is not contaminated by the environment, otherwise it defeats the purpose.
Horizontal surfaces, where particles most easily settle. This includes the tops of equipment, rolling carts, tables, or even the inside shelves of personnel airlocks.
And finally, perhaps the most important point: human contact surfaces. This includes door handles, benches in the airlocks, computer stations, tablets, and the surfaces used to fill out batch records, among others.
And how do you determine the sampling frequency?
For an EMPQ, there is no regulatory requirement defining the frequency. It is up to the manufacturer to propose a coherent strategy that can provide a complete picture of how well the cleanroom environment is controlled.
What we often see in practice is 3 days of sampling in static conditions and 3 days in dynamic conditions. But other strategies are also possible.
How do you determine the type of sampling to perform: active air, settle plates, or surface monitoring?
In this case, it’s defined by regulation. Annex 1 of the EU GMP provides two tables with the acceptance limits for Grades A, B, C, and D.
The first table covers non-viable particles, and the second covers viable particles. For viable particles, three methods must be applied, shown in three separate columns: surface monitoring, active air sampling, and settle plates.
Can you explain the difference between these three types of sampling?
Active air sampling measures instantaneous airborne contamination. Settle plates measure microbiological fallout over time.
And surface sampling, generally using contact plates, evaluates overall hygiene and, more importantly, the effectiveness of surface cleaning and disinfection procedures. This method is especially relevant when focusing on human-contact points.
Why do we combine these three methods in an environmental monitoring strategy?
We combine them because they do not target the same phenomena. Active air sampling focuses on the ambient air.
- For surface monitoring, it’s simple: you target the surfaces.
- For settle plates, you target the air that deposits onto horizontal surfaces.
And contrary to a common belief we often hear in companies “Settle plates outside Grade A are useless” that is actually false.
There are two important biases with surface sampling:
The recovery rate is limited: if a surface contains 100 Micrococcus luteus, a contact plate will not give you 100 CFU. The gap is even greater if you don’t use an applicator to standardize the pressure.
Contaminants released in an airlock do not necessarily settle where you sample. For example, when you remove your gown in an airlock, you release particles and microorganisms that can stay airborne for a long time, but they will not necessarily deposit on the bench where you take the contact plate.
Hence the importance of combining methods to obtain a complete picture of contamination.
Which regulations or standards (Annex 1, ISO, etc.) guide the definition of sampling points and methods?
Of course. First, we have the Good Manufacturing Practices, especially Annex 1 for the manufacture of sterile medicinal products. Even if a company does not produce sterile or low-bioburden products, it still needs to refer to Annex 1, because it is the only document that clearly describes the expectations for cleanroom classification.
And finally, to determine the minimum sampling volume based on the room’s square meters, we rely on ISO 17141 for viable particles and ISO 14644 for non-viable particles.
Adaptation of the monitoring plan throughout the cleanroom’s lifecycle
Are the routine sampling points necessarily the same as those defined during the EMPQ?
They can be the same, but they can also be different. If they are modified, whether a sampling point is moved, or the sampling volume is increased or reduced, it is possible, but it must be justified.
The justification may come from a new risk assessment or from a trend review that leads to such a change.
Can these points be modified during the cleanroom’s lifecycle?
Of course, I would even say that they should be. The monitoring program must be a living system.
How do you justify these changes?
This is done through periodic trend reviews and the analysis of critical steps (and for those who have a Contamination Control Strategy, this can be linked to it).
However, if a new risk is identified (for example, through the introduction of a new process, a deviation, or a change control), then additional sampling points may be added.
If a sampling point is consistently negative and is also non-critical, then yes, it can be removed.
Does this happen often in practice?
It depends on the company’s activity. For a company that has been running the same process for 10 years and will continue doing so for the next 10 years, there is usually no strong reason to change anything, unless a trend excursion occurs.
But for a company opening a new site, and especially for a CDMO with many different processes, then yes, monitoring points may change. This is even more likely because a brand-new site has very limited historical data on its environmental trends.
What role do microbiological trend analyses play in environmental monitoring?
Their role is clearly central: they act as a real “excursion radar,” providing:
- quantitative information, through control charts,
- qualitative information, on the types of microorganisms, their recurrence, and especially the evolution of the site’s endogenous flora,
- as well as spatial information between different cleanroom grades, and seasonal trends.
Trend analysis plays three major roles:
A documentation role: during audits and inspections, it demonstrates that the cleanrooms are maintained in a state of control.
A preventive role: detecting drifts before they become non-conformities and helping identify root causes.
A management role: making the right decisions based on the results and adjusting the monitoring program when needed.
And how do you use this information in practice to adjust or evolve the monitoring strategy?
You can adjust the sampling frequency, the location of sampling points, and the level of decontamination.
For frequency, for example, if an area has shown stable values for more than 12 months, always below alert limits, you might reduce the sampling frequency or the number of sampling points (as long as it is justified through a risk assessment). The idea is obviously not to fall into the mindset of “well, nothing ever grows here, so I’ll stop monitoring.”
Conversely, if an area shows regular fluctuations close to the limits, you can increase the frequency to better monitor stability.
For the positioning of sampling points, you look at spatial trending, which shows where excursions occur most often. This may lead to moving some points or adding new ones.
For the level of decontamination, you review the recurrence of certain organisms in the trends. If needed, you may consider a different biocide.
If there is an increase in skin flora microorganisms, you focus on personnel training and proper handwashing practices.
If recurring fungal isolates appear, you verify the effectiveness of the disinfectants and sporicides used. If they are effective, you then look at the decontamination method: is it applied correctly? Is it appropriate?
“We observed a tenfold increase, and in some cases much more, in mold recovery at 20–25 °C. (for the first incubation temperature)”
Choice of incubation temperatures for environmental monitoring
What are the most common incubation temperatures and sequences?
I’ve seen several practices. Some use 30–35 °C followed by 20–25 °C, others 20–26 °C followed by 30–35 °C, and others a single incubation temperature, such as 28 °C.
How do you justify the choice of these conditions?
In my opinion, it must be based on a risk assessment, supported by a study of the site’s endogenous flora.
The manufacturer must characterize the endogenous flora of the cleanrooms and adapt its microbiological monitoring program accordingly.
If an area shows a high presence of molds, you’re not going to start with incubation at 30–35 °C, otherwise you would completely miss them.
In one of my past experiences, we performed an active air sample in a material airlock between the warehouse and the cleanroom. When comparing incubation starting at 20–25 °C versus 30–35 °C, we observed a tenfold increase, and in some cases much more, in mold recovery at 20–25 °C.
So all incubation conditions, time, temperature, and media, must be justified through an analysis of the endogenous flora.
And can these parameters also evolve over time, depending on the results or on the type of product being manufactured?
Of course, they must evolve. A site’s endogenous flora is constantly changing, depending on:
- the arrival of new cleanroom operators,
- the introduction of new materials and equipment,
- seasonality,
- or even the type of crops being grown around the site.
For example, one year there may be corn, and the next year wheat… you don’t bring the same microorganisms into the cleanroom. And we should not forget that humans are basically buses for microorganisms: when we spend time outdoors, our skin and hair “pick up” whatever is in the environment.
You can even imagine other factors that influence the evolution of endogenous flora: air pollution, weather conditions, construction work on the building, and so on.
In reality, defining incubation parameters in 2025 does not guarantee at all that they will still be appropriate in 2026.
“The first trap is to perform monitoring without any real rationale — just to tick a box and show that you’re doing it during an audit or inspection.”
Main pitfalls to avoid when setting up environmental monitoring
What do you think are the main pitfalls to avoid when implementing an environmental monitoring program?
The first pitfall is to perform monitoring without any rationale, just to tick a box and show that you’re doing it during an audit or inspection. You need to rely on a real risk assessment and a solid understanding of your process.
The second pitfall would be to overload the sampling plan without having the operational capacity to actually perform and analyze all the samples. The risk is ending up with a sampling team overwhelmed by the workload and by data management. It’s better to have fewer sampling points, but ones that are intelligently chosen and properly analyzed. Once again, risk assessment is extremely important for prioritization.
And a third pitfall: thinking that a cleanroom is just a floor, walls, and a ceiling. In reality, there are many elements interacting within a cleanroom:
- the water loops,
- the gas networks,
- the personnel,
- incoming materials,
- the safety showers,
- the drains, etc.
And all of these can introduce contaminants at different moments (or sometimes none at all)… but they must be taken into account in the monitoring strategy.
A big thank you to Cédric for taking the time to share his expertise and for shedding so much light on EMPQ and environmental monitoring in pharmaceutical cleanrooms.