The expanded application of most probable number to the quantitative evaluation of extremely low microbial count.

This paper is about the evaluation of the extremely low microbial counts from the field by expanding the most probable number (MPN) methodology when the data follow Poisson distribution in order to achieve more accurate estimation with limited number of data. Several data sets with extremely low microbial counts from pharmaceutical clean room and laboratory clean bench monitoring were found with good Poisson distribution fitness by chi2 test (P < 0.05), and the MPN calculations were conducted for these data by using the Halvorson and Ziegler equation. The MPN values are generally larger than the arithmetic average, indicating a higher sensitivity of the data assay. This approach is justified because Poisson distribution is the mathematical background of the MPN procedure, with the Halvorson and Ziegler equation as its foundation. It is considered that the MPN methodology has a potential application for quality control in the extremely low level microbial counting environment in the clean rooms levels ISO class 7 or above. Further studies on the precision of this method and development of a sampling plan based on careful mathematical analysis will help to refine the approach.