This article is written in order to address some areas of misunderstanding within the medical manufacturing industry regarding measurements of microbial contamination in cleanroom theaters as promoted in some documents.^1,2 Amongst these issues of confusion are the use of settle plates, and establishing alert and action limits for routine biological monitoring of cleanroom operations.

1 L. Soltis; “Developing an Environmental Monitoring Sampling Plan,”
A2C2 Magazine, (January, 2001).2 US Pharmacopeia, “Microbiological Evaluation of Clean Rooms and Other Controlled Environments,” <1116> (USP 25, 2002).3 “Fundamentals of a Microbiological Environmental Monitoring Program,” Journal of Parenteral Science & Technology, Technical report No. 13, Revised, (2001).4 L. Pasteur. “Mémoire sur les corpuscles organisÈs qui existent leans l’atmosphére; examen de la doctrine de générations spontanées,” Ann. Sci. Nat. (1861).5 W. Whyte. “In Support of Settle Plates,” PDA Journal of Pharmaceutical Science & Technology, Vol. 50, No. 4, (July-August 1996).6 N. Burris, A. Elfe. “Validating Biocleanrooms for Terminally Sterilized Medical Devices,” Microcontamination, (March, 1993).7 M. P. Russell, J. A. Goldsmith, L. Philips. “Some Factors Affecting the Efficiency of Settle Plates,” J. Hosp. Infect., (1984).8 PDA Comments: “USP on Microbiological Evaluation of Clean Rooms and Other Controlled Environments <1116>, PDA Journal of Pharmaceutical Science & Technology, Vol. 51, No. 6, (1997).9 B. Ljungqvist, B. Reinmuller. “Airborne Viable Particles and Total Number of Airborne Particles: Comparative Studies of Active Air Sampling,” PDA Journal of Pharmaceutical Science & Technology, Vol.54, No. 2, (March-April, 2000).10 B. Ljungqvist, B. Reinmuller, Berit. “ The Biotest RCS Air Samplers in Unidirectional Flow,” Journal of Pharmaceutical Science & Technology, Vol. 48, No. 1, (January-February, 1994).11 Dr. J. E. Akers. “Comparative Clean Room GMP’s,” PDA Validation of Pharmaceuticals and Biopharmaceuticals, (February 1993).  J. Akers, J. Agallaco. “Environmental Monitoring: Myths and Misapplications,” PDA Journal of Pharmaceutical Science & Technology, Vol. 55, No. 3, (May-June, 2001).

Biologically Active Particles & Settle Plates

The issue of the use of “settle plates” is identified in respected references³ and the passive capture of settling microbes represents one of the oldest measurement systems for collecting airborne microbes.⁴ One article stated, “Because of air turbulence and random particle settling, settling plates are not quite accurate.”¹ This statement is inaccurate. Settle plates measure exactly that phenomenon for which they are intended—the random settling of biologically active particles (BAPs).

As will be shown, the relevance of settle plates corresponds directly to the potential impact of BAPs on the process and product, while the relevance of active samplers is uncertain. In this respect settle plates are certainly quantitative for that phenomenon, and one need not be concerned about trying to correlate settle plate recovered values to active sampling measurements of airborne BAPs.

Even in respected references this appears to be an area of confusion when it is suggested that settle plates should be considered qualitative in nature.² Though we disagree with such statements, the words of Whyte⁵ capture the scientific response most succinctly, “The following are the most common criticisms made of the use of settle plates and the replies that may be made to them. Settle plates do not measure the concentrations of microorganisms in the air.” As a statement this is correct. However, it cannot be used as a criticism. Settle plates do not measure the number of microorganisms in the air but measure the number of microorganisms settling from air onto a known surface area in a known time.

Settle plate measurements may be more relevant to assessing potential distribution of BAPs on product than either surface or air sampling. Regardless of which method or combination of methods are employed, one conclusion is certain; each of the three methods measure different environmental contamination attributes and there is no direct correlation between any two as a general rule though good correlation has been reported in some specific cases.⁶ Although the sensitivity of settle plates is often criticized, Russell⁷ has shown settle plates can be left out up to eight hours without loss of sensitivity.

Interpretation of Sampling Methods & Techniques

Validation of either contact recovery or air sampling methods is a point often avoided by the industry, with good reason. The resulting counts are always technique and equipment specific.⁸ That is, there is no method of determining the true level of BAPs; therefore, there is no way to determine the correlation of the results of such sampling methods to the true value. Using two different types of air samplers in the same environment will give different results because each imparts their own bias into the measurement.

Affecting biases would include: sample flow rates, orifice size and related angle to the air flow, particle size distribution in the sampled air, potential disruption of the local air flow to produce a non-statistical sample,⁹ as well as the potential effects of dehydration on the media and collected BAP organisms. For some air samplers, calibration of sampled air volumes is an issue. In addition, air sampling measures some particles that would otherwise remain entrained in the air stream and be exhausted out of the environment. In such cases some particles measured by active sampling methods would never become actual surface contaminants on products, components or equipment. The elimination efficiency of the dynamic environmental system is dependent on air speeds, eddy currents, particle sizes and charges. An environmental microbial validation plan can only determine the values obtained by some specific method of sampling and culturing and then correlate these results to the activities, conditions and product bioburden values during the evaluation period.

It is similarly difficult to validate the recovery efficiency of contact plates because the efficiency is affected by: sampled surface material and associated electrostatic charges; smoothness of the surface; residual antimicrobials; wetness of the media and the nature; and charge and size of the BAPs. Since these variables will change from one environment, surface and equipment set to another, a single validation cannot be universally applied. In addition, attempting to use a surface inoculated with one organism type to validate recovery efficiency is not representative of the environmental organisms present, the nature of the particulates carrying the biological entities or the method of deposition and subsequent adherence of organisms to the surface.

The use of active samplers is often implemented without due consideration of their potential impact, both on the sample itself and on the process and products. Active samplers should NEVER be hand held during the sampling process. The effects of personnel on airflow patterns may be extensive. The disruptive nature of air influx and efflux during sampling has been studied by Ljungqvist and Reinmuller for RCS samplers.^9,10 Also, the more rapid the sampling rate the greater the expected disruption and potential impact on both sample and process. The combination of disruptive influences by operator and the sampling instrument during hand held sampling may result in a sample of the operator more than of the process. Since these disruptions are certain, they should be minimized, particularly where they could adversely affect “the microbiological safety of the process and the product.”¹⁰ The least disruptive samplers are those where the sampled air discharge outlets are connected to an exhaust line which can be directed to a return vent or outside the clean environment. At least, the discharge should be directed away from the sampled area as well as from critical environmental areas, considering the process and products.

Establishing Alert & Action Limits

Companies often implement alert and action limits based only on the biological sampling data. This may result in frequent alert and action responses because the data gathering sequence was too short to provide the range of reasonably expected results. Seasonal variation, for example, frequently has significant statistical impact for ISO cleanroom levels 7-9. At the same time environmental limits have been exceeded, actual product test data may be far below product limits. In this scenario, companies create a reaction/response cycle based on arbitrarily imposed limits from original environmental sample results for the process validation. The escape from this cycle is to float limits upward based on correlation of new environmental data with product data collected at the same time. This approach allows a two-way street within which manufacturers can operate.

When system controls are inadequate to provide safe and efficacious products, then the controls must be tightened. Conversely, when system controls far exceed those required for adequate control of a production area, the identified limits may be loosened, based on actual correlated test data.

It should be noted that processes during which the product is wiped with 70% IPA (or another suitable disinfectant), especially just prior to final packaging, often produce product bioburden levels little related to the production environment bioburden levels. In these situations, if high product levels are detected, microbial sleuth work is necessary to determine the source of the offending microbial population.³ It is imperative not to lower environmental bioburden limits without corroborating product data, as the environment is often not the causative source. As Akers has noted, “I believe that the suggestion of numerical limits without validated standard methods is scientifically unsound. Furthermore, I do not believe such limits, however well meaning, provide any additional safety for the health care consumer”.¹¹ We echo the call for good science and pragmatic approaches¹² when establishing environmental bioburden limits for manufacturing operations. Manufacturers should recognize that microbiological limits are equipment and technique specific and the limits may be increased or decreased based on supporting data.

References

12