A method for determining the size of internal pilot studies

Area of Science:

• Biostatistics
• Clinical Trial Design
• Statistical Methods

Background:

• Accurate sample size calculation is crucial for clinical trials comparing population means.
• Low precision in variance estimates for the main efficacy variable complicates sample size determination.
• Internal pilot studies offer a potential solution by utilizing early trial data.

Purpose of the Study:

• To propose a method for determining the size of internal pilot studies in clinical trials.
• To ensure the internal pilot study size is maximized without exceeding the optimal planned study size.
• To discuss the advantages and limitations of the proposed method.

Main Methods:

• The study suggests a method for calculating the appropriate sample size for an internal pilot study.
• This method aims to balance the need for a sufficiently large pilot study with the constraint of not exceeding the final planned sample size.
• The approach focuses on refining the variance estimate using data from the initial phase of the trial.

Main Results:

• The proposed method provides a structured approach to defining the size of internal pilot studies.
• It allows for adaptive sample size adjustments based on early data, improving efficiency.
• The method ensures the pilot study size is optimally determined within predefined limits.

Conclusions:

• Internal pilot studies are valuable for refining sample size estimates in clinical trials with uncertain variance.
• The proposed method offers a statistically sound way to determine internal pilot study size.
• This approach can lead to more efficient and reliable clinical trial designs.