Using Confidence Distributions in Final and Interim Analyses for Single-Arm Studies or Platform Trials Consisting of...

Area of Science:

• Statistics
• Clinical Trial Design
• Pharmacometrics

Background:

• Confidence distributions are gaining traction as a frequentist alternative to Bayesian posterior distributions due to their simplicity.
• Single-arm and platform trials are increasingly utilized in rare diseases, oncology, and pediatric drug development for proof-of-concept and pivotal evidence.
• These trials often employ two-stage designs with sequential or continuous monitoring, analyzed via frequentist or Bayesian methods.

Purpose of the Study:

• To describe analysis strategies using confidence distributions for single-arm and platform trials.
• To focus on binary endpoints and define decision rules for final and interim analyses.
• To derive operating characteristics exactly, without simulation, for these trial designs.

Main Methods:

• Development of analysis strategies based on confidence distributions for single-arm and platform trials.
• Application to binary endpoints, defining decision rules for interim and final analyses.
• Utilization of predictive probabilities for interim decision rules, contrasting with stochastic curtailment's conditional probabilities.

Main Results:

• Exact derivation of operating characteristics for confidence distribution-based analyses is demonstrated, eliminating the need for simulations.
• A novel approach using predictive probabilities for interim decision rules in clinical trials is presented.
• The methodology is shown to be applicable to various complex trial designs including platform, basket, and umbrella trials.

Conclusions:

• Confidence distributions provide a viable and exact frequentist approach for analyzing single-arm and platform trials.
• The proposed method, using predictive probabilities, simplifies the definition and analysis of interim decision rules.
• This framework enhances the statistical rigor and efficiency of drug development studies, particularly in specialized therapeutic areas.