An efficient algorithm to determine the optimal two-stage randomized multinomial designs in oncology clinical trials

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Oncology Research

Background:

• Phase II oncology trials often use two-stage randomized multinomial designs considering response rate (RR) and early progression rate (EPR).
• Determining optimal design parameters for these trials can be computationally intensive, limiting practical application.
• Sun et al. (2009) proposed a design but faced computational challenges.

Discussion:

• This paper presents an efficient algorithm to identify optimal two-stage randomized multinomial designs.
• The algorithm employs an approximation method and other techniques to substantially reduce computational intensity.
• This addresses the computational burden of previous optimal design parameter determination.

Key Insights:

• The proposed algorithm achieves over a 90% reduction in computation time.
• The approximation error is acceptably low, ensuring reliable results.
• This enhances the feasibility and usability of optimal two-stage multinomial designs.

Outlook:

• The efficient algorithm may increase the adoption of optimal two-stage multinomial designs in clinical practice.
• It potentially enables the extension of these designs to more complex clinical trial scenarios.
• Further research could explore its application in adaptive trial designs.