臨床試驗中，所謂的調適設計是指根據不同治療組別之反應結果再去分配下一位病患使用其中一種治療的方法。因為符合道德上的需求，所以近年來此方法在臨床實驗設計中被廣泛討論。針對存活時間資料，目前大部分的文獻多為探討在兩組治療組別下的依反應結果調整的調適設計，較少針對多組治療組別討論。本研究考慮針對反應變數為存活時間在服從指數分配的情形下，依反應結果調整的設計且設定滿足某些特定的目標下，加以考慮利用最佳化調適比例分配病患接受不同的治療，推導出多組治療組別相對之最佳化調適比例。本論文將藉由模擬研究呈現多組治療情形下，和傳統的平衡設計互相比較，其結果發現依反應結果調整的調適設計能夠將更多的病患分配至接受風險較低的治療。最後，本論文會藉由一組Byar和Green(1980)的攝護腺癌臨床資料說明在多組治療組別下，說明依反應結果調整的調適設計如何應用在臨床上。

In clinical trials, response-adaptive randomization becomes popular recently. Because it can change the allocation probabilities sequentially, with the main goal of assigning more patients to the better treatment. However, most relevant references focus on two-armed clinical trials especially that response outcome is survival. In this paper, we derived two optimal allocation proportions for multi-armed survival trials. And the doubly adaptive biased coin design is applied to develop a randomization procedure. In order to compare the efficacy of treatments, we use multiple Wald test to control the Type I error rate not higher than significant level α. Through simulation study, we verified our method can allocate patients to most efficacious treatment than the popular balance designs. We also illustrate our allocation procedure by redesigning a four-armed prostate cancer trial and make comparisons with balance designs.

[1] Byar, DP., Green, SB. (1980). The choice of treatment for cancer patients based on covariate information-application to prostate cancer. Bulletin Du Cancer, 67(4), 477-90.
[2] Bartlett, RH., Roloff, DW., Cornell, RG., Andrews, AF., Dillon, PW., Zwischenberger, JB. (1985). Extracorporeal circulation in neonatal respiratory failure: a prospective randomized study. Pediatrics, 76(4), 479-487.
[3] Hu, F., Zhang, L. (2004). Asymptotic properties of doubly adaptive biased coin designs for multi-treatment clinical trials. The Annals of Statistics, 32, 268-301.
[4] Zhang, L., Rosenberger, W. (2007). Response-adaptive randomization for survival trials: the parametric approach. Journal of the Royal Statistical Society. Series C (Applied Statistics), 56(2), 153-165.
[5] Hu, F., Rosenberger, W. (2003). Optimality, variability, power: evaluating response-adaptive randomization procedures for treatment comparisons. Journal of the American Statistical Association, 98, 671-678.
[6] Zhang, L., Rosenberger, W. (2006). Response-adaptive randomization for clinical trials with continuous outcomes. Biometrics, 62(2), 562-569.
[7] Cheung, YK., Inoue, LYT., Wathen, JK., Thall, PF. (2006). Continuous bayesian adaptive randomization based on event times with covariates. Statistics in Medicine, 25, 55-70.
[8] Zhu, H., Hu, F. (2009). Implementing optimal allocation for sequential continuous responses with multiple treatments. Journal of Statistical Planning and Inference, 139, 2420-2430.
[9] Eisele, JR., Woodroofe, MB. (1995). Central limit theorems for doubly adaptive biased coin designs. The Annals of Statistics, 23(1), 234-254.
[10] Lawless, JF. (2002). Statistical models and methods for lifetime data. New York: Wiley.
[11] Rosenberger, W., Stallard, N., Ivanova, A., Harper, CN., Ricks, ML. (2001). Optimal adaptive designs for binary response trials. Biometrics, 57, 909-913.
[12] Rosenberger, W., Seshaiyer, P. (1997). Adaptive survival trials. Journal of Biopharmaceutical Statistics, 7, 617-624.
[13] Eisele, JR. (1994). The doubly adaptive biased coin design for sequential clinical trials. Journal of Statistical Planning and Inference, 38, 249-261.
[14] Dunnett, CW. (1955). A multiple comparison procedure for comparing several treatments with a control. Journal of the Royal Statistical Society, 50(272), 1096-1121.
[15] Harrell, FE. (2001). Regression modeling strategies. New York: Springer-Verlag.