For label claims or indications, especially in the US, prospectively designed clinical trials are usually required. However, when robust data are available and indicate a nearly unmistakable performance of the product of interest, the post hoc analysis should be seriously considered for the label claim or indication purpose. In this article, post hoc, Bayesian, and stand-alone new trial analyses are discussed and compared so that under such situations, a decision-making process could be facilitated.
US Food and Drug Administration. Draft Guidance for Industry, Clinical Investigators, and Food and Drug Administration Staff: Design Considerations for Pivotal Clinical Investigations for Medical Devices. Silver Spring, Maryland: FDA; 2011.
2.
FarbA.When is a randomized clinical trial appropriate vs. a historical control vs. a performance goal?Paper presented at: FDA/CDRH, Optimizing Clinical Trial Design for the Development of Pediatric Cardiovascular Devices; September 30, 2010; San Francisco, California.
3.
European Medicines Agency. Benefit-Risk Methodology Project. Work Package 2 Report: Applicability of Current Tools and Processes for Regulatory Benefit-Risk Assessment. London: EMA; 2010.
4.
RosenbaumPR. Observational Studies. Springer Series in Statistics. New York, NY: Springer-Verlag; 2010.
5.
WangJ. Aperspective on the connection between Bayesian and frequentist approaches. In: Joint Statistical Meetings Proceedings. Biopharmaceutical Section. Alexandria, Virginia: American Statistical Association; 2010:2153–2158.
6.
US Food and Drug Administration. Guidance for the Use of Bayesian Statistics in Medical Device Clinical Trials. Silver Spring, Maryland: FDA; 2010.
7.
HobbsBPSargentDJCarlinBP. Commensurate priors for incorporating historical information in clinical trials using general and generalized linear models. Bayesian Anal. 2012;7(2):1–36.
