Exact two-stage designs for phase II activity trials with rank-based endpoints

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Statistical Design

Background:

• Phase II clinical trials often involve limited knowledge of experimental treatments and resource constraints, necessitating efficient designs.
• Existing two-stage designs primarily focus on binary endpoints using exact calculations or continuous endpoints using approximations.
• There is a need for robust statistical designs for continuous endpoints in phase II trials that are both accurate and practical.

Purpose of the Study:

• To present novel exact two-stage Mann-Whitney designs specifically for two-arm randomized clinical trials.
• To provide practical tools for trial planning by offering decision rules based on various assumed scenarios.
• To address the limitations of existing designs for continuous endpoints in phase II settings.

Main Methods:

• Development of exact two-stage designs utilizing the Mann-Whitney U test for continuous endpoints.
• Application of these designs within the framework of two-arm randomized clinical trials.
• Generation of comprehensive tables detailing decision rules for trial planning under diverse assumptions.

Main Results:

• The proposed designs provide exact probability calculations for continuous endpoints in two-stage phase II trials.
• Tables of decision rules are presented, facilitating informed planning and execution of clinical trials.
• These methods enhance the statistical rigor and efficiency of phase II trial designs.

Conclusions:

• Exact two-stage Mann-Whitney designs offer a statistically sound and practical approach for phase II clinical trials with continuous outcomes.
• The presented decision rules aid researchers in planning trials effectively, considering resource limitations and ethical considerations.
• This work contributes to the advancement of clinical trial methodology, particularly for studies evaluating new treatments with continuous endpoints.