Designs for adding a treatment arm to an ongoing clinical trial

Area of Science:

• Clinical trial design
• Biostatistics
• Drug development

Background:

• Clinical trials often involve multiple treatments at various development stages.
• Adding a new experimental treatment arm mid-trial presents statistical challenges, particularly regarding multiple comparisons.
• Current practices often compare new arms only to concurrently randomized controls.

Purpose of the Study:

• To propose an adaptive trial design for adding a new treatment arm.
• To control the family-wise error rate (FWER) while maintaining marginal power.
• To explore optimal allocation strategies for enhanced study power.

Main Methods:

• An adaptive design is proposed to adjust sample size upon adding a new treatment arm.
• The design aims to control FWER in the strong sense.
• Optimal allocation is investigated to maximize the probability of detecting all superior treatments.

Main Results:

• Maintaining marginal power and FWER control necessitates a sample size increase when adding an arm.
• In some scenarios, separate trials may be more efficient than a single trial with an added arm.
• Optimal allocation after adding an arm increases overall study power compared to equal allocation.

Conclusions:

• Adding treatment arms requires careful statistical planning to manage error rates and power.
• Adaptive designs and optimal allocation are crucial for efficient multi-arm trials.
• The choice between adding an arm or conducting separate trials depends on specific design characteristics and efficiency goals.