Take Your NI Trial to the Next Level
Reflecting the vast research on noninferiority (NI) designs from the past 15 years, Noninferiority Testing in Clinical Trials: Issues and Challenges explains how to choose the NI margin as a small fraction of the therapeutic effect of the active control in a clinical trial. Requiring no prior knowledge of NI testing, the book is easily accessible to both statisticians and nonstatisticians involved in drug development.
With over 20 years of experience in this area, the author introduces the basic elements of the NI trials one at a time in a logical order. He discusses issues with estimating the effect size based on historical placebo control trials of the active control. The book covers fundamental concepts related to NI trials, such as assay sensitivity, constancy assumption, discounting, and preservation. It also describes patient populations, three-arm trials, and the equivalence of three or more groups.
Table of Contents
Equivalence and Noninferiority Testing
One-Sided and Two-Sided Hypotheses
Equivalence Margin δ
Choice of δ
Ethical Issues and Declaration of Helsinki
Conventional Approach versus Role-Reversal Approach
Notations and Noninferiority Hypotheses
Choice of Noninferiority Margin for the Mean Difference
Choice of ε and Interpretation
How Do You Know the Effect Size?
Constancy Assumption versus Assay Sensitivity
Choice of Noninferiority Margin for the Mean Ratio and Hazard Ratio
Why Mean Ratio?
Mean Ratio for Continuous Endpoints
Hazard Ratio for Survival Endpoints
Noninferiority Hypotheses with Binary Endpoints
Noninferiority Hypothesis Based on the Difference or Ratio of Two Proportions
NI Hypothesis Based on the Odds Ratio
Two Statistical Approaches for Testing the Noninferiority Hypothesis
One-Sided versus Two-Sided
Fixed-Margin Method versus Synthesis Method
Preservation versus Discounting
Control of Type I Error Rates
Switching between Superiority and Noninferiority
Statistical Issues with Simultaneous Testing
Conditional Type I Error Rate in Superiority Testing
Discussion and Concluding Remarks
Multiple Historical Studies and Meta-Analysis
Meta-Analysis and Systematic Review
Constancy Assumption in the Context of Meta-Analysis
Three Treatment Groups
Two Test Treatments versus an Active Control
Equivalence of Three Test Groups
Guidelines in the Anti-Infective Area with a Binary Endpoint
FDA Draft Guidance
Intention-to-Treat versus Per-Protocol
Analyses of Randomized Control Trials
Intention-to-Treat Analysis versus Per-Protocol Analysis in Superiority Trials
Intention-to-Treat Analysis versus Per-Protocol Analysis in Noninferiority Trials
Implementation of an Intention-to-Treat Analysis in Randomized Control Trials
CBER Criteria in the Assessment of New Thrombotics in June 1999
Issues and Challenges
Advances in Noninferiority Trials
Current Controversial Issues
Issues and Challenges
References appear at the end of each chapter.
"The author gives interesting examples and discusses all issues critically with respect to existing literature. Furthermore, striking and convincing motivation is given as well as discussions at the ends of most chapters. Many diagrams and graphs clearly illustrate the respective content and statistical issues. The book allows readers from different fields to understand the concept of NI testing as well as issues and challenges related to NI testing."
"This book is intended for statisticians and non-statisticians working in the area of drug development, and it is written to be easily understood by a broad audience without any prior knowledge of non-inferiority testing. It covers important aspects of non-inferiority testing, including onesided and two-sided hypotheses, equivalence/non-inferiority margins, fixed-margin and the synthesis methods, controversial issues with switching between superiority and non-inferiority testing, tests with binary, continuous and time-to-event endpoints, multiple historical studies and meta-analysis, regulatory guidances, and issues and challenges in these tests."
~Lili Zhao, University of Michigan