Statistical inference for response adaptive randomization procedures with adjusted optimal allocation proportions

ABSTRACT Seamless phase II/III clinical trials have attracted increasing attention recently. They mainly use Bayesian response adaptive randomization (RAR) designs. There has been little research into seamless clinical trials using frequentist RAR designs because of the difficulty in performing valid statistical inference following this procedure. The well-designed frequentist RAR designs can target theoretically optimal allocation proportions, and they have explicit asymptotic results. In this paper, we study the asymptotic properties of frequentist RAR designs with adjusted target allocation proportions, and investigate statistical inference for this procedure. The properties of the proposed design provide an important theoretical foundation for advanced seamless clinical trials. Our numerical studies demonstrate that the design is ethical and efficient.

[1]  D. Berry.  Bayesian Statistics and the Efficiency and Ethics of Clinical Trials , 2004 .

[2]  L. J. Wei,et al.  The Randomized Play-the-Winner Rule in Medical Trials , 1978 .

[3]  William F. Rosenberger,et al.  Optimality, Variability, Power , 2003 .

[4]  M. Zelen,et al.  Play the Winner Rule and the Controlled Clinical Trial , 1969 .

[5]  C. Assaid,et al.  The Theory of Response-Adaptive Randomization in Clinical Trials , 2007 .

[6]  L. Hayre.  Two-population sequential tests with three hypotheses , 1979 .

[7]  KyungMann Kim.  Group Sequential Methods with Applications to Clinical Trials , 2001 .

[8]  William F. Rosenberger,et al.  Implementing Optimal Allocation in Sequential Binary Response Experiments , 2007 .

[9]  D. Berry.  Adaptive clinical trials in oncology , 2012, Nature Reviews Clinical Oncology.

[10]  G. Parmigiani,et al.  Bayesian adaptive randomized trial design for patients with recurrent glioblastoma. , 2012, Journal of clinical oncology : official journal of the American Society of Clinical Oncology.

[11]  W. R. Thompson.  ON THE LIKELIHOOD THAT ONE UNKNOWN PROBABILITY EXCEEDS ANOTHER IN VIEW OF THE EVIDENCE OF TWO SAMPLES , 1933 .

[12]  H. Robbins.  Some aspects of the sequential design of experiments , 1952 .

[13]  N Stallard,et al.  Optimal Adaptive Designs for Binary Response Trials , 2001, Biometrics.

[14]  Feifang Hu,et al.  Asymptotic properties of doubly adaptive biased coin designs for multitreatment clinical trials , 2003 .

[15]  Jianhua Hu,et al.  Optimal biased coins for two-arm clinical trials , 2008 .

[16]  J. Matthews,et al.  Randomization in Clinical Trials: Theory and Practice; , 2003 .

[17]  P F Thall,et al.  A two-stage design for choosing among several experimental treatments and a control in clinical trials. , 1989, Biometrics.

[18]  W F Rosenberger,et al.  A COMPARISON OF URN DESIGNS FOR RANDOMIZED CLINICAL TRIALS OF K > 2 TREATMENTS , 2000, Journal of biopharmaceutical statistics.

[19]  Jing Ning,et al.  Using short‐term response information to facilitate adaptive randomization for survival clinical trials , 2009, Statistics in medicine.

[20]  Peter F Thall,et al.  Seamlessly Expanding a Randomized Phase II Trial to Phase III , 2002, Biometrics.