Bayesian sample-size determination methods considering both worthwhileness and unpromisingness for exploratory...

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmacoeconomics

Background:

• Exploratory clinical trials are crucial for go/no-go decisions on experimental treatments.
• Adequate sample size is essential for reliable assessment of treatment efficacy or lack thereof.
• Current methods may not sufficiently address the probabilistic nature of exploratory trial outcomes.

Purpose of the Study:

• To propose and evaluate three new sample-size determination methods for randomized exploratory clinical trials.
• To ensure that sample sizes provide convincing evidence for treatment worthwhileness or unpromisingness.
• To integrate posterior probabilities with predefined efficacy and inefficacy criteria.

Main Methods:

• Development of three novel sample-size determination methodologies.
• Utilizing posterior probabilities derived from efficacy and inefficacy criteria.
• Conducting simulation studies, including numerical investigations, for validation.

Main Results:

• The proposed methods demonstrated a high probability of correctly classifying treatments.
• Experimental treatments with superior response probabilities were declared worthwhile.
• Treatments with inferior response probabilities were reliably declared unpromising.

Conclusions:

• The new sample-size methods offer a robust approach for exploratory clinical trials.
• These methods enhance the reliability of go/no-go decisions in early-phase drug development.
• Accurate sample size determination is key to efficient and informative clinical research.