Adding Subjects or Adding Measurements in Repeated Measurement Studies Under Financial Constraints

Area of Science:

• Biostatistics
• Clinical Trial Design
• Health Economics

Background:

• Clinical trial budget constraints necessitate careful design choices to maximize statistical power.
• Repeated measures studies require balancing the number of subjects and measurements per subject.
• Financial limitations are compounded by factors like missing data and complex correlation structures.

Purpose of the Study:

• To develop an optimal design strategy for repeated measures clinical trials under budget constraints.
• To maximize statistical power for a given financial investment.
• To account for complexities such as missing data and various correlation structures.

Main Methods:

• Proposed an approach combining Generalized Estimating Equations (GEE) for slope coefficient estimation with cost constraint optimization.
• Derived an analytical solution for optimal design with no missing data and compound symmetric correlation.
• Employed numerical search for optimal design when missing data or other correlation structures are present.
• Conducted an extensive simulation study to evaluate design impacts.

Main Results:

• The optimal balance between the number of subjects and repeated measurements is sensitive to cost ratios and correlation structures.
• Missing data and dropout rates significantly influence the optimal design and overall study power.
• The proposed GEE-based approach effectively identifies cost-efficient designs across various scenarios.

Conclusions:

• A systematic approach combining GEE and cost constraints can optimize clinical trial design for repeated measures studies.
• Careful consideration of potential missing data and correlation patterns is crucial for efficient trial planning.
• The findings offer practical guidance for investigators managing budget limitations in clinical research.