Probability monitoring procedures for sample size determination

Area of Science:

• Clinical Research
• Biostatistics
• Drug Development

Background:

• Power analysis is standard for sample size calculation in clinical research to detect meaningful differences.
• Traditional methods are often infeasible for clinical trials with very low incidence rates or rare diseases due to immense sample size requirements and limited patient populations.

Purpose of the Study:

• To propose an innovative sample size determination method for clinical trials where standard power analysis is not feasible.
• To introduce a probability monitoring procedure for controlling the likelihood of crossing critical safety and/or efficacy boundaries.

Main Methods:

• A novel probability monitoring procedure is proposed for sample size determination.
• This method focuses on controlling the probability of crossing predefined safety and/or efficacy thresholds.
• An adaptive probability monitoring procedure is suggested for multi-stage adaptive trial designs in rare disease development.

Main Results:

• The proposed method offers a viable alternative to traditional power analysis for sample size determination in specific clinical trial contexts.
• It provides statistical assurance by managing the probability of critical boundary crossings.
• Adaptive procedures can enhance flexibility in rare disease clinical development.

Conclusions:

• The probability monitoring procedure offers a practical approach to sample size determination when traditional power analysis is not feasible.
• This method is particularly relevant for rare disease drug development and trials with low incidence rates.
• Adaptive probability monitoring can further optimize sample size and trial conduct in complex clinical research scenarios.