Using the partitioning principle to adaptively design dose-response studies.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmacology

Background:

• Dose-response studies are crucial for determining optimal drug dosages.
• Traditional designs may be inefficient in identifying the minimum effective dose.
• Controlling statistical error rates is paramount in clinical research.

Purpose of the Study:

• To propose an adaptive two-stage design for dose-response studies.
• To efficiently identify the minimum effective dose (MED).
• To maintain strong control over the familywise error rate (FWER).

Main Methods:

• A two-stage adaptive design based on the partition testing principle.
• Stage 1: Examination of a wide dose range against a placebo.
• Interim analysis for potential design modifications, including termination of ineffective/unsafe doses.
• Stage 2: Further investigation of selected doses.
• Inference based on pre-chosen conditional error functions.

Main Results:

• The proposed adaptive design allows for efficient dose selection.
• Ineffective or unsafe doses can be identified and stopped early.
• The familywise error rate is strongly controlled throughout the study.
• Conditional error functions provide a flexible framework for inference.

Conclusions:

• The adaptive two-stage design offers an efficient and statistically sound approach for dose-response studies.
• This methodology aids in identifying the minimum effective dose while managing resources effectively.
• The use of conditional error functions enhances the adaptability and robustness of the design.