Review of alert and action level setting for microbiological cleanroom data

 

Scientific data is concerned with measuring and hence data, whether that is qualitative or quantitative. In pharmaceutical microbiology, this could be a number of cells or colony forming units, a series of growth or no growth results; and incidences of microorganisms. Gathering such data allows for trending and enables control to be achieved.

 

With microbial numbers, it is a regulatory expectation that alert and action levels be set. Alert and action levels are not specifications - they are ‘snap-shot’ indicators of potential adverse or upward trends, or out-of-control situations. Alert and action levels are used to detect shifts from the norm and to indicate if an individual result or process is potentially out-of-control. Therefore, what is important is the data pattern.

 

The standard approach is to set alert and action levels based on a set of historical data. Understanding past data enables for a more accurate assessment of current data in relation to the performance of a sample, system or environment. Furthermore, the use of this information ensures that the levels applied relate in some form to past data rather than being based on an arbitrary figure. Setting alert and action levels is hampered by data sets being distributed unusually, with the data containing many zeros (and so skewed to the left when a histogram is produced). This type of quality attribute is typically called a zero-inflated variable. Furthermore, there will often be occasional high values leading to a relatively high dispersal. The data is not always predictable and will contain a relatively high level of variability. One way to overcome this is through the use of the percentile cut-off approach.

 

Sandle, T. (2023) Review of alert and action level setting for microbiological cleanroom data, RSSL Life Science Insights, at: https://www.rssl.com/insights/life-science-pharmaceuticals/review-of-alert-and-action-level-setting-for-microbiological-cleanroom-data/

 

Posted by Dr. Tim Sandle, Pharmaceutical Microbiology Resources (http://www.pharmamicroresources.com/)