Information-adaptive clinical trials with selective recruitment and binary outcomes

Area of Science:

• Clinical Trial Design
• Biostatistics
• Medical Informatics

Background:

• Traditional clinical trials may not optimally utilize participant information.
• Participant selection and treatment allocation can significantly impact trial efficiency and statistical power.

Purpose of the Study:

• To define and evaluate methods for quantifying statistical information in clinical trial recruitment.
• To assess the impact of selective recruitment and information-adaptive designs on statistical power and trial efficiency.

Main Methods:

• Developed and evaluated four distinct methods for quantifying statistical information.
• Utilized experimental data and numerical simulations to compare selective recruitment designs with traditional randomized designs.
• Investigated information-adaptive allocation of recruits to treatment arms.

Main Results:

• Selective recruitment designs demonstrated a substantial increase in statistical power compared to randomized designs.
• Information-adaptive allocation in non-selective trials also improved statistical power.
• Selective recruitment offers potential for trials with fewer participants, yielding economic and ethical benefits.

Conclusions:

• Selective recruitment and information-adaptive designs significantly enhance statistical power in clinical trials.
• These advanced designs offer a pathway to more efficient, cost-effective, and ethically sound clinical research.
• Optimizing participant selection and allocation is crucial for maximizing information gain and trial success.