PT  - JOURNAL ARTICLE
AU  - Bernhard Schmelzer
AU  - André Mischo
AU  - Franz Innerbichler
TI  - Statistically significant versus practically relevant trend in stability data
AID  - 10.5731/pdajpst.2020.012336
DP  - 2021 Jan 01
TA  - PDA Journal of Pharmaceutical Science and Technology
PG  - pdajpst.2020.012336
4099  - http://journal.pda.org/content/early/2021/02/19/pdajpst.2020.012336.short
4100  - http://journal.pda.org/content/early/2021/02/19/pdajpst.2020.012336.full
AB  - How to decide whether a statistically significant trend is of practical relevance? In the context of stability data of pharmaceuticals, this publication provides a way forward to use different measures of (method) variability to compare to the observed changes over time. A panel of analytical experts assessed whether statistical significant trends were of practical relevance or not. For different types of assessing variability, recommendations for decision criteria were derived that best match these assessments i.e. finding a suited balance between not detecting a relevant trend and between flagging a trend wrongly as relevant. For this purpose, more than 60 data sets from Biosimilar projects of Sandoz/Novartis were leveraged. Hence, this paper provides a scientific way for an assessment whether a statistically significant trend is of practical relevance or not and a case study is presented and discussed.