Sequential designs with small samples: Evaluation and recommendations for normal responses.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Rare Disease Research

Background:

• Sequential monitoring methods are increasingly recommended for rare disease clinical trials due to reduced sample size requirements.
• Existing sequential methods often rely on large-sample approximations, which can inflate Type I error rates in small patient populations typical of rare diseases.
• Traditional clinical trial design paradigms may be unsuitable for rare diseases where patient availability is limited.

Purpose of the Study:

• To evaluate the performance of sequential designs with small to moderate sample sizes for normally distributed outcomes.
• To demonstrate the necessity of simple corrections to critical boundaries for accurate sequential testing in rare disease trials.
• To propose a method for selecting optimal sequential designs when a maximum sample size is a constraint.

Main Methods:

• Operational characteristics of sequential designs were assessed using normally distributed outcomes.
• Simulations or analyses were performed for very small to moderate sample sizes.
• Corrections to critical boundaries were evaluated.
• A method for optimal sequential design selection was developed considering maximum sample size and prior treatment effect beliefs.

Main Results:

• Sequential methods relying on asymptotic assumptions can lead to inflated Type I error rates in small sample sizes.
• Simple corrections to critical boundaries are necessary to maintain the desired Type I error probability.
• The proposed method allows for the selection of an optimal sequential design under a maximum sample size constraint.

Conclusions:

• Adjustments to sequential monitoring methods are crucial for rare disease clinical trials with limited patient numbers.
• Correcting critical boundaries is essential for reliable results when using sequential methods in small populations.
• The developed approach facilitates efficient clinical trial design in rare diseases by optimizing sequential strategies within sample size limitations.