在臨床試驗的發展中，適應性無縫設計包含兩個階段，學習階段（二期階段）選擇劑量或試驗組，確認階段（三期階段）確認所選劑量或試驗組的治療效果。由於在一個方案中結合傳統的二三期試驗，以至於有可能可以縮短新藥開發所需要的時間而備受製藥公司關注。本研究提出一個除了在二期階段結束時，基於短期終點選擇一個試驗組外，並增加一個新的試驗組在三期階段的無縫設計。在最終分析時，利用封閉檢定法控制由於劑量選擇造成可能的總體型一錯誤率膨脹，另外使用逆常態加權組合函數結合兩階段基於入組受試者晚期終點所得到p值以進行統計推論。本研究提出依序檢定與同時檢定兩種檢定策略，依序檢定首先檢定二期階段選出的試驗組，若達顯著則再檢定新增的試驗組；同時檢定則同時檢定兩個試驗組。本研究透過大量的模擬，以驗證此設計在不同情境下能控制總體型一錯誤率並達到足夠的檢定力。;In the development of clinical trial design, adaptive seamless design contains learning stage and confirmatory stage corresponding to conventional phase Ⅱ/Ⅲ trial respectively. Typically, experimental groups as doses of test drug or treatment groups compared to an active control are evaluated at the learning stage, based on the results of the learning stage, the efficacy of selected dose or treatment group is to be confirmed at the confirmatory stage. As adaptive seamless design combines conventional phase Ⅱ/Ⅲ trial into a trial design, it might shorten time needed to develop new drugs, which has drawn considerable attention from researchers and pharmaceutical companies. We proposed a seamless design which is able to select an experimental group at the end of learning stage based on short-term outcome and to add a new experimental group into confirmatory stage. The closed testing procedure and weighted inverse normal combination function are used to control the inflation of overall type I error rate due to dose selection and p-values from two stage subjects at the final analysis. We proposed sequentially and simultaneously testing strategy for testing the new experimental group. Sequentially testing strategy tests the selected experimental group first, if significant then test the new experimental group. Simultaneously testing strategy tests the selected and new experimental groups at the same time. Numerical simulation has been conducted to verify the type I error rate and power under several scenarios.
