N.54 – Application of MPN in microbial air environmental tests


When the air is aspirated through the aspirating head of the air samplers indicated in the USP 1116 (Slit to agar samplers, Sieve impactors, Centrifugal samplers, Surface Air System) and micro-organisms impact upon the agar surface there is a statistical probability that the micro-organisms may land on top of each other and be visually interpreted as one single colony (CFU).


Non Controlled Environments

This situation happens when the air “bio-burden” is quite high and typical of “non-controlled environment” or “open spaces”.

To overcome this possibility, the Most Probable Number (MPN) table were created first by Andersen and later by Janet Macher.

With “TRIO.BAS”, MPN Table, the count correction is not used until 16 colonies are identified (the application only starts with more than 16 counted colonies).


Clean room / RABS / Isolators

The use of the MPN table is not practically applied in Clean Rooms.


– The different grades in Clean Rooms

-Clean room (pharmaceutical)

Grade A

Air Dynamic < 1 CFU / 1000 litres of air

Air Sedimentation < 1 CFU / plate / 4 hours

Surface < 1 CFU / 25 square centimetres

Operator < 1 CFU / glove

Grade B

Air Dinamic 10 CFU / 1000 litres of air

Air Sedimentation 5 CFU / plate / 4 hours

Surface 5 CFU / 25 square centimetres

Operator 5 CFU / glove

Grade C

Air Dinamic 100 CFU / 1000 litres of air

Air Sedimentation 50 CFU / plate / 4 hours

Surface 25 CFU / 25 square centimetres

Grade D

Air Dinamic 200 CFU / 1000 litres of air

Air Sedimentation 100 CFU / plate / 4 hours

Surface 50 C FU/ 25 square centimetres

MPN application: Yes or NO?

If you look at the requested specifications of different Grades, the following considerations can be taken into account, using an active agar microbiological air sampler:

The pharma plants have generally an expectation of relatively low counts in a carefully controlled environment.

Accuracy is paramount and excursion can jeopardize batch release.


As a practical matter, alert levels and action levels are so low in an ISO 5 (Class 100) environment that the MPN will seldom be reached and, if so, will be by definition a variation from tolerance.

Hence there is no call for extra complication of MPN.


In more contaminated environments like ISO 7 and ISO 8, the number of counted colonies can exceed the level (100 or 200 CFU / 1000 litres of air) and invoke the MPN table application adjustments. If this case is sporadic, simply take the visual count.

When TNTC (Too Numerous To Count) plates are encountered routinely, the “simplest” and “best” approach is to reduce the sampling volume (e.g.: from 1.000 litres to 200 litres) which will perforce reduce the CFU counts and reach a range through experience where the count may be below 16 CFU.

Statistical confidence in the results is increased by (i) taking multiple samples at the lower volume range or (ii) adopting the dual heads “DUO or TRIO air sampler, increasing the “N”.



What is important is not whether MPN is used with its inherent limitations, but rather that the operator is consistent in the used method. As long as the data is consistently treated the same, the trend is what is most important. Alert and action levels will be adjusted annually to reflect the number reported, be they absolute or relative. Where data is used for comparison, it is often best to standardize on the simplest practical method, rather than seeking perfection with its inherent danger of misapplication.



USP 1116