[[alternative]]Curtailed Two-Stage Design for Comparing Two Binomial Proportions in Phase Ii Clinical Trials

系統編號: PA10207-0393計畫編號: NSC102-2118-M343-001執行機關: 南華大學自然生物科技學系研究期間: 10208~10307[[abstract]]在新藥研發的過程中，受試者通常是依序進入臨床試驗中，並逐一記錄其療效反應。所以，當受試者的招募不易時，要累積到足夠的受試者人數，往往需要很長的時間。針對單組試驗設計，Chi 和Chen (2008) 發展出一種縮減兩階段設計來加速臨床試驗進行。在單組試驗設計中，由於並無對照組，因此在比較上是藉由參與此臨床試驗之受試者的療效反應率與歷史對照組之療效反應率進行比較，因此受試者間可能存在較大的異質性，使得在兩組反應率差的估計上產生偏誤，增加研究結果的不確定性，所以隨機對照試驗設計越來越受到重視。然而，此縮減兩階段設計並無法直接應用在隨機對照試驗中。因此，在本研究將根據縮減設計的改概念發展縮減隨機對照二階段設計，根據早期試驗之結果做成決策，加快臨床試驗的進行。[[abstract]]In phase II clinical trial, patients are recruited sequentially and consequently the duration of the trial will become very long if the accrual rate is very low. Chi and Chen (2008) proposed a curtailed two-stage design which used in a single-arm trial to speed up the clinical trial process. In the singlearm design, it compares the response rate of the new drug with a preselected fixed target response rate estimated from a historical study. The patient populations may be quite heterogeneous for the two populations. Therefore, the potential bias may arise in comparing two response rates. To achieve the patient comparability, a randomized controlled study is often applied. However, the randomized controlled study is hard to use the curtailed design immediately. Therefore, in this paper, the concept of curtailed designs will be applied to a randomized controlled trial to shorten the drug development process