FDA Opens for Comment Revised ICH Guideline on Statistical Principles for Clinical Trials
Posted 30 October 2017 | By
The US Food and Drug Administration (FDA) on Monday opened for comment a newly revised International Council for Harmonisation (ICH) guideline intended to better align the choice of statistical methods with questions of regulatory importance, and to improve the reliability of decisions about and representations of the effects of medical products.
The 31-page draft guidance, according to FDA, clarifies, updates and extends the earlier ICH guidance, entitled "E9 Statistical Principles for Clinical Trials," in two main areas.
"Concerning estimands, it provides a framework for discussion of how the aims of a trial relate to the proposed statistical analysis. Concerning sensitivity analysis, it discusses how to use additional analyses to address concerns about the validity of assumptions underlying the main analysis," the agency said in a Federal Register notice.
The draft guidance looks to align the choice of statistical methods with the goals of a clinical trial, to communicate the rationale for such choices to FDA, and on using sensitivity analysis, to characterize the robustness of the conclusions to plausible deviations from the assumptions of the main analysis.
In July 2017, the ICH Assembly endorsed the E9(R1) draft, which is a product of ICH’s Efficacy Expert Working Group. Comments submitted on this draft will be considered by FDA and the Efficacy Expert Working Group.
E9(R1) Statistical Principles for Clinical Trials: Addendum: Estimands and Sensitivity Analysis in Clinical Trials