Abstract
A validated method is used to measure quality attributes of manufactured lots to support the production of active ingredient or drug product for a regulatory filing or product release. Various regulatory guidelines provide expectations on how to validate a method, including accuracy and precision. Validation studies are not typically designed to have the size or scope needed to fully characterize a method. We can and should, however, capitalize on the historical data that is generated from method development, transfer, and clinical trial lot release to facilitate the characterization. Using prior information from these sources affords a more confident characterization of the method’s accuracy and precision. This article focuses on Bayesian methods for using the stream of available method data. Considerations are given for prior distributions, including methods to control the influence of prior data. Two case studies are presented to illustrate the continuous validation approach. This is a joint effort of the DIA Bayesian Scientific Working Group and the American Statistical Association Biopharmaceutical Nonclinical Working Group.
Funding
The author(s) reported there is no funding associated with the work featured in this article.