公衛所 Faculty Seminar
時間：110年4月26日(一) 12:10 – 13:10
(Time: 12:10 PM, 26^th April)
地點：醫學二館221室
(Location: Rm221, Medical Building)
講者：許根寧老師(藥明康德新藥開發有限公司首席生物統計師)
(Speaker: Dr. Ken-Ning Hsu, WuXi AppTec, Shanghai)
Title：Seamless Phase II/III Design Using Early Outcome for Dose Selection Followed Adding an Additional Treatment Method of Selected Dose
In recent years, adaptive seamless phase II/III designs have been used in clinical trials, in which the data at an interim analysis is used for treatment or dose selection. Typically, more than one dose of test drug compared to an active control are evaluated at the phase II, i.e., the stage I. Based on the results of the phase II, additional subjects may be randomized into the selected doses and the active control at the phase III, i.e. the stage II. All subjects are followed until the measurement of the primary endpoint is observed. Suppose in a clinical trial all subjects are randomized into two doses of test drug and an active control drug at the stage I, and one dose of test drug is selected based on the results of the stage I for continuation into stage 2. At the phase III, the additional subjects would be randomized into two different treatment methods of the selected dose and the active control. According to the result of previous studies, the dose selection is based on early outcome in order to save time. In this talk, I am going to introduce the statistical inference. The numerical simulation has been conducted to investigate the type I error rate and power under several scenarios, the results will be presented.