在藥物開發過程中，第II期臨床試驗其中一個重要的目的是進行新藥效力的初步評估。然而，由於在第一期臨床試驗中，新藥的最大耐受劑量的估計可能並不十分準確，使得在第二期臨床試驗中所使用的劑量並不恰當，導致出現過度的毒性反應。因此，針對單組設計的第二期臨床試驗，Conaway和Petroni (1995) 與Jin (2007) 等學者紛紛提出可同時監控新藥療效反應率與毒性反應率的二階段設計。然而，就我們所知，針對雙組設計的第二期臨床試驗，目前並無可同時比較兩組療效反應率與毒性反應率的二階段隨機對照設計。因此，本計畫擬針對在同時考慮新藥有效性與安全性的情況下，提出一個合適的二階段隨機對照設計，在虛無假設成立下，盡可能的將型一誤差控制在最小。此外，為加快臨床試驗的進行，本計畫亦考慮將縮減抽樣程序應用於二階段隨機對照設計中，發展出一個適用於兩相依二元反應變項之縮減二階段隨機對照設計。 In a drug development process, an important goal of phase II clinical trial is to evaluate the initial efficacy of a new drug. However, the maximum tolerated dose of the new drug might not always be accurately estimated in the phase I clinical trial. As a result, the recommended dose could lead to excessive toxicity in the subsequent phase II clinical trial. For single-arm trials, Conaway and Petroni (1995), and Jin (2007), among others have developed several two-stage designs to simultaneously evaluate the efficacy and safety of new drugs. To the best of our knowledge, there are no available designs for bivariate binary endpoints in a two-arm trial. Therefore, this project proposes a two-stage randomized design with two dependent binary endpoints. In the proposed design, the type I error rate is controlled over the entire null region to be as small as possible. In addition, in order to speed up the trial process, the curtailed sampling procedure will also be applied to the two-stage randomized design to develop a curtailed two-stage randomized design with two dependent binary endpoints.
