In Part I of this article series, the regression control chart method for identifying out-of-trend data in pharmaceutical stability studies is investigated, and an improved approach is suggested.
kentoh/shutterstock.com
The regulation of pharmaceutical stability studies still lacks universally accepted techniques regarding the identification of out-of-trend data. Three methods have been suggested for identifying out-of-trend data in pharmaceutical stability studies: the regression control chart, the by-time-point method, and the slope control chart. In Part I of this article series, the regression control chart method is investigated, and an improved approach is suggested. The method is illustrated using realistic data. In Part II, the by-time-point method and the multivariate control chart method will be discussed.
Click here to view a PDF of this article.Peer-Review Article
Submitted: March 22, 2017
Accepted: June 27, 2017
Máté Mihalovits is a PhD student, mihalovits@mail.bme.hu; and Sándor Kemény* is an emeritus professor, Tel.: +36 309936307, kemeny@mail.bme.hu, both at Budapest University of Technology and Economics, Faculty of Chemical Technology and Biotechnology, Department of Chemical and Environmental Process Engineering Hungary, 1111. Budapest, Műegyetem rakpart 3.
* To whom all correspondence should be addressed.
Pharmaceutical Technology
Vol. 41, No. 11
November 2017
Pages: 46–53
When referring to this article, please cite it as M. Mihalovits and S. Kemény, "Methods for Identifying Out-of-Trend Data in Analysis of Stability Measurements–Part I: Regression Control Chart," Pharmaceutical Technology 41 (11) 2017.
A Novel, Enhanced, and Sustainable Approach to Audit Trail Review
July 4th 2025Eli Lilly and Company developed an innovative and sustainable approach to audit trail review (ATR) aimed at reducing the ATR burden while adhering to regulatory expectations and data integrity (DI) principles. The process has transformed employees' understanding of ATR and complemented the DI by design approach, leading to better system designs that meet expected controls and reduce non-value-added data reviews.