Strategy to equivalence testing for development and scale up of biopharmaceutical downstream processes

Gamma, E-Beam and X-ray Irradiations on PE/EVOH/PE Multilayer Film: An Industrial Point of View Regarding the Impact on Mechanical Properties

“…δ is a hypothetical value such that if the absolute value of the observed difference is no more than δ, there is a strong probability of concluding that the two datasets represent equivalent results [40]. There are different methods to calculate the equivalence acceptance criterion [36], such as the 3-sigma rule (the one used in this paper), Limentani's EAC, Limentani modified, Cohen's standardized difference-effect sizes-and tolerance interval.…”

“…Today, a simple equivalence test first introduced by Schuirmann et al [37] is accepted, for example, as standard for bioequivalence assessment [38], namely, the two one-sided t-test (TOST). In contrast to the two-sample t-test, the null hypothesis of the TOST states that the two means are not equivalent [36]. The impact of the null hypothesis is that in the case of small sample sizes and/or poor precision (large variance) in one or both groups, equivalence is rejected, resulting in low numbers of false-positive test results [39].…”

“…Typically, hypothesis testing is used as a data analysis technique to evaluate the comparability of data. Our approach here aims to be as close as possible to the industrial conditions of irradiations, and two datasets are comparable if their difference in means is of no practical and statistical significance such that it is accepted as nearly zero [36]. Today, a simple equivalence test first introduced by Schuirmann et al [37] is accepted, for example, as standard for bioequivalence assessment [38], namely, the two one-sided t-test (TOST).…”