- 學位論文
以條件模式為基礎的一期臨床試驗之二藥物劑量訂定方法 ：Thall et al. 方法的另一種作法
Conditional Modeling-Based Approach for Two Dimensional Dose Finding in Phase I Trials : An Alternative Approach of Thall et al.’s Method
若您是本文的作者，可授權文章由華藝線上圖書館中協助推廣。
並列摘要
The primary purpose of a phase I clinical trial is to determine the safety dose of a new drug for later trials use. Many statistical methods have been proposed for the determination of dose for a single drug. In statistical sense it can be regarded as a two-dimensional dose-toxicity curve fitting problem. In many trials, however, investigators are more interested in finding the doses for multiple agents. When there are multiple agents, statistical problem becomes more complicated due to the interaction among agents, and the monotonicity between toxicity and dose can no longer be assumed as that in the case of a single agent. For a two-agent trial, we have to deal with the three-dimensional curve fitting problem. A previous method used a joint-probability regression model to undergo the model-setting for toxicity response curve with two agents. Hence, we modify this method by dividing the model into two conditional models. Each conditional model describes the toxicity reaction of an agent conditional on the other agent. The advantage of our method is to avoid complexity in estimation but still retain the capability of the original method in coping with interaction between two agents.
參考文獻
行政院衛生署 (1998). 藥品臨床試驗一般基準（台北）.
Arbuck, S. G. (1996). Workshop on Phase I study design. Annals of Oncology 7, 567–573.
Casella, G. and Berger, R. L. (2002). Statistical Inference, 2nd edition. Pacific Grove, CA: Duxbury.
Dent, S. F. and Eisenhauer, E. A. (1996). Phase I trial design: Are new methodologies being put into practice? Annals of Oncology 7, 561–566.
Dobson, A. J. (2002). An Introduction to Generalized Linear Models, 2nd edition. Boca Raton, Florida: Chapman & Hall/CRC.