Confirmatory analysis for phase III population pharmacokinetics
Article first published online: 29 DEC 2009
Copyright © 2009 John Wiley & Sons, Ltd.
Volume 10, Issue 1, pages 14–26, January/February 2011
How to Cite
Hu, C., Zhang, J. and Zhou, H. (2011), Confirmatory analysis for phase III population pharmacokinetics. Pharmaceut. Statist., 10: 14–26. doi: 10.1002/pst.403
- Issue published online: 29 DEC 2009
- Article first published online: 29 DEC 2009
- Manuscript Accepted: 13 AUG 2009
- Manuscript Revised: 24 JUL 2009
- Manuscript Received: 17 FEB 2009
- model selection;
- covariate modeling;
- exploratory analysis;
- stepwise search;
Population pharmacokinetics (POPPK) has many important uses at various stages of drug development and approval. At the phase III stage, one of the major uses of POPPK is to identify covariate influences on human pharmacokinetics, which is important for potential dose adjustment and drug labeling. One common analysis approach is nonlinear mixed-effect modeling, which typically involves time-consuming extensive search for best fits among a large number of possible models. We propose that the analysis goal can be better achieved with a more standard confirmatory statistical analysis approach, which uses a prespecified primary analysis and additional sensitivity analyses. We illustrate this approach using a phase III study data set and compare the result with that calculated using the common exploratory approach. We argue that the confirmatory approach not only substantially reduces analysis time but also yields more accurate and interpretable results. Some aspects of this confirmatory approach may also be extended to data analysis in earlier stages of clinical drug development, i.e. phase II and phase I. Copyright © 2009 John Wiley & Sons, Ltd.