在第二期臨床試驗中，兩治療組隨機試驗的主要目是要瞭解新療法的臨床有效性與標準療 法相比是否具有較佳的療效。除此之外，基於倫理的考量，設計上通常以兩階段的方式進行臨 床試驗，以避免給太多的病人無療效的藥。然而，在一個前導性研究中，由於研究人員對於新、 舊治療之研究資訊往往相當缺乏，導致研究人員在決定新、舊治療之療效反應率差的時候，常 常感到困擾。有鑑於此，本計畫擬發展一個調適性二階段設計，其特點是允許研究人員可以根 據第一階段所觀察到之有效反應人數的差異，彈性的選擇療效反應率差。假如所觀察到呈有效 反應之人數，差異夠大，則研究人員傾向於選擇一個樂觀的療效反應率差與較高的檢定力；反 之，則選擇一個保守的療效反應率差與較低的檢定力。並在給定的顯著水準與檢定力下，根據 大中取小與最佳化兩種準則，找出最合適的試驗設計參數。 The main purpose of a two-arm randomized trial in phase II clinical trials is to determine whether the clinical efficacy of treatment therapy is superior to the standard therapy. In addition, for ethical considerations, the design of two-stage clinical trial is commonly used to avoid giving patient an ineffective therapy. However, in a pilot study, the investigators often face uncertainty in deciding the clinically significant difference in response rates for the two therapies, especially in lack of the information about the two therapies in literature. In view of this, this plan will propose an adaptive two-stage design, in which the clinically significant difference in response rates can be flexibly decided based on the gap in the numbers of responses observed in the first stage. If the gap is large enough, the investigators will tend to choose an optimistic clinically difference with a higher power; whereas, a conservative clinically difference will be chosen with a lower power. Moreover, for the pre-specified significant level and power, the suitable design parameter will be derived based on Simon’s minimax and optimal criteria.
