Proper inference from Simon's two-stage designs.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmaceutical Research

Background:

• Simon's two-stage designs are widely used in phase II clinical trials.
• Current inference methods often disregard the adaptive nature and actual sample size of these designs.
• This oversight can lead to inaccurate statistical reporting.

Purpose of the Study:

• To develop and present a method for statistical inference in Simon's two-stage designs that accounts for the actual sampling plan.
• To compare the proposed inference method with conventional approaches.
• To address the discrepancies arising from ignoring adaptiveness and actual sample sizes.

Main Methods:

• A literature review of inference procedures for Simon's designs was conducted.
• A novel method for statistical inference was developed, incorporating both planned and actual sample sizes.
• P-values, point estimates, and confidence intervals were calculated using both conventional and proposed methods.

Main Results:

• Inference procedures commonly ignore the adaptive sampling plan and actual sample sizes in Simon's designs.
• The proposed method provides inferences adjusted for the actual sample size.
• While conventional and preferred methods often yield similar P-values and confidence intervals, conventional inference may conflict with hypothesis testing outcomes.

Conclusions:

• Accurate statistical inference in adaptive clinical trial designs requires accounting for the actual sample size.
• The proposed method enhances the reliability of statistical reporting in phase II trials using Simon's designs.
• Ignoring adaptiveness can lead to inconsistencies in hypothesis testing and statistical conclusions.