Implementing unequal randomization in clinical trials with heterogeneous treatment costs

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Health Economics

Background:

• Equal randomization is standard in clinical trials.
• Heterogeneous variances, variable costs, and maximizing participant benefit may necessitate unequal allocation proportions.

Purpose of the Study:

• Investigate optimal allocation designs for parallel group clinical trials.
• Minimize study costs under statistical efficiency constraints when comparing investigational treatments to a control.

Main Methods:

• Theoretical investigation of allocation designs.
• Exploration of restricted randomization procedures for practical implementation.
• Application of population-based and randomization-based approaches for statistical inference.

Main Results:

• Equal allocation designs can be suboptimal.
• Unequal allocation designs offer higher statistical power for the same budget or reduced cost for the same power.
• Unequal randomization designs are cost-efficient and practically implementable.

Conclusions:

• The choice of target allocation, randomization procedure, and statistical analysis is crucial for valid, powerful, and robust clinical trial results.
• Unequal randomization designs offer significant advantages in specific clinical trial settings.