Adaptive sample size modification in clinical trials: start small then ask for more?

Area of Science:

• Clinical Trial Design
• Biostatistics
• Medical Research Methodology

Background:

• Sample size re-estimation is crucial in clinical trials, especially with delayed patient response.
• Existing methods, like Mehta and Pocock's 'promising zone,' have potential application pitfalls.

Purpose of the Study:

• To analyze the limitations of current sample size re-estimation methods in clinical trials.
• To propose improved sample size adjustment rules for enhanced trial efficiency and power.

Main Methods:

• Analysis of existing sample size re-estimation methodologies, including those by Mehta and Pocock.
• Development of new sample size rules based on the principle of maximizing the benefit of additional observations.

Main Results:

• Identified inefficiencies in the 'promising zone' approach, where maximum power gains may lie outside defined regions.
• Demonstrated that moderate sample size increases over a wider range of outcomes are more efficient than large increases for narrow outcome ranges.

Conclusions:

• The framework for sample size re-estimation is valuable but requires careful application to avoid pitfalls.
• Proposed sample size rules offer a more efficient approach to adaptive clinical trial design, particularly for delayed responses.