Wednesday, December 28, 2016

New Paper on Using Statistics When Validating Rapid Microbiological Methods

From 2010 to 2013, European Pharmaceutical Review published a very successful series on rapid microbiological methods (RMM) that included such topics as how to demonstrate a rapid method is equivalent to a conventional or pharmacopoeial method. PDA Technical Report #33, the recently revised USP chapter <1223> and the proposed changes to Ph. Eur. Chapter 5.1.6 all recommend the use of statistical analysis when validating rapid methods and establishing equivalence. However, many microbiologists find the concepts of statistical analysis and hypothesis testing intimidating and, in some cases, unapproachable. Once and for all, it was time to demystify statistics and provide a framework for embracing the mathematical models required when validating novel RMM technologies. 

To accomplish this, I have teamed up with two of the industry’s leading experts on statistical analysis and the validation of rapid microbiological methods: Edwin R. van den Heuvel, PhD (Eindhoven University of Technology) and David Roesti, PhD (Novartis Pharmaceuticals).

Our paper, The Role of Statistical Analysis in Validating Rapid Microbiological Methods, has recently been published in this month’s European Pharmaceutical Review. Clicking on the publication title (above) will give you online access to the paper (this is free to view but you may need to sign up for access). Alternatively, please copy and past the following address into your web browser:

This comprehensive article discusses the concepts of statistics and hypothesis testing, including:
  • How to formulate a null and alternate hypotheses as it relates to rapid method validation 
  • Identifing the appropriate statistical test
  • How to calculate a p-value
  • Comparing the p-value with a chosen alpha level
Next we, dive into equivalence and non-inferiority tests, which we believe is the preferred strategy for demonstrating a rapid method is as good as and/or no worse than the existing microbiological method you wish to replace. The article specifies how to select lower and upper equivalence levels and also provides case studies on how to utilize these statistical tools.


Post a Comment