Doctoral Candidate Junyao Wang defends her dissertation on "Adaptive Randomization in Sequential Multiple Assignment Randomized Trials" 

Advisor: Abdus Wahed, PhD, Department of Biostatistics

Committee Members: 

• Stewart Anderson, PhD, Emeritus Department of Biostatistics
• Joyce Chang, PhD, Department of Medicine
• Ying Ding, PhD, Department of Biostatistics
• Yu Cheng, PhD, Department of Statistics

ABSTRACT:

Sequential multiple assignment randomized trials (SMARTs) are systematic and efficient media for comparing dynamic treatment regimes (DTRs), where each patient is involved in multiple stages of treatment with the randomization at each stage depending on the patient's previous treatment history and interim outcomes. They have become increasingly popular in the field of precision medicine. Generally, in SMARTs, patients are randomized equally to ethically acceptable treatment options regardless of how effective those treatments were during the previous stages, which results in some undesirable consequences in practice, such as low recruitment, less retention, and lower treatment adherence. In clinical trials, response-adaptive randomization (RAR) has been proposed to alleviate such concerns. In this dissertation, we introduce the between-patient RAR technique into the design of SMART to benefit more patients with more efficacy regimes, further to increase recruitment and retention in a SMART.

In the first part of the dissertation, we propose a response-adaptive SMART (RA-SMART) design where the randomization probabilities are imbalanced in favor of more promising treatments based on the accumulated information on treatment efficacy from previous patients and stages. The operating characteristics of the RA-SMART design relative to SMART design, including the consistency and efficiency of estimated response rate under each DTR, the power of identifying the optimal DTR, and the number of patients treated with the optimal and the worst DTRs, are assessed through extensive simulation studies. Finally, some practical suggestions on RA-SMARTs are discussed in conclusion.

In the second part of the dissertation, we propose an optimal response-adaptive SMART (ORA-SMART) design for minimizing the total number of failures in the trial without sacrificing the power for identifying the optimal DTR embedded in the trial. We provide an algorithm to solve for optimal randomization probabilities at stage II, applying the Majorization-Minimization  method. The proposed ORA-SMART is compared to the regular SMART with equal randomization probabilities at stage II under various scenarios through simulation. We conclude with a discussion on potential implications and future research.

Public health significance: Sequential multiple assignment randomized trials (SMARTs) have been widely used in clinical research of treatment sequencing in many disease areas such as behavioral and mental health and oncology. The designs proposed in this dissertation will help alleviate ethical concerns about treating patients with less effective treatments, and in the sequel, will improve public health.