Dear Sir or Madame,
In the last few years, classical hypothesis testing for determination of assay and sample suitability in biological assays has been the target of much criticism. Due to advancements in measurement techniques and assay preparation workflows, most modern assays are capable of obtaining replicate observations with very low variability. It is well known that for such assays the hypothesis tests, or F-tests, on significance of regression, non-parallelism and non-linearity will become overly sensitive. Using these tests not only risks discarding perfectly fine assays and causing costly and cumbersome re-runs, but they also create a misleading incentive where imprecise work yields improved test results. Luckily, hypothesis testing isn't the only approach for determining the suitability of your samples or your whole assay.
Equivalence Testing: The modern approach to testing your assay
Since the disadvantages of hypothesis testing are well known, Chapter <1032> of the US Pharmacopoeia recommends to use equivalence testing for determining your sample similarity. However, overall adaption of equivalence testing is still relatively low. Partially, this might be explained by the fact that this method isn't as ready-to-use as hypothesis testing. Hypothesis testing only requires choosing an adequate confidence level. In contrast, to perform effective equivalence testing it is not only necessary to determine suitable measures of similarity but also to determine the margins against which these measures are tested. The USP states:
"The challenge in implementing equivalence testing is setting appropriate equivalence bounds for the nonsimilarity measures"
Usually, these measures of similarity and their margins are developed from historical data, but manually collecting and processing all the relevant parameters is cumbersome and prone to errors. PLA 3.0 already stores your historical data in the documents on your database. With PLA 3.0 you can start developing your margins right away.
Using your assay archive for margin development
The Equivalence Margin Development document of the Biological Assay package in PLA 3.0 is capable of extracting and processing the relevant information from the Quantitative Response Assays already stored in your database. With a few clicks, you can select your development assays, develop margins, and define, verify and visualize different testing strategies. You can also export your new margins as a test system definition to include in your future QRAs. So there's no need for manual transfer from Equivalence Margin Development to assay either.
Stay tuned for next week's newsletter, where we will give a short overview of the five principal components of the Equivalence Margin Development document, how to use them and how they work together during margin development.