Inclusion of non-inferiority analysis in superiority-based clinical trials with single-arm, two-stage Simon's design.

摘要: Abstract Introduction Non-inferiority (NI) analysis is not usually considered in the early phases of clinical development. In some negative phase II trials, a post-hoc NI justified additional III trials that were successful. However, risk false positive achievements was controlled these analyses. We propose to preplan analyses superiority-based Simon's two-stage designs control type I and error rates. Methods Simulations have been proposed assess errors rates with this method. A total 12,768 design constructed based on different assumptions rejection response probability, desired errors, margins. P-value calculated stochastic ordering using Uniformly Minimum Variance Unbiased Estimator. Type simulated Monte Carlo The agreement between values analyzed Bland-Altman plots. Results observed same level from both superiority which allowed. Different examples has explain utility Conclusion Inclusion single-arm may be useful for weighing factors such as safety, pharmacokinetics, pharmacodynamic, biomarker data while assessing efficacy. Implementation strategy can achieved through simple adaptations existing one-arm trials.