光華講壇——社會(huì)名流與企業(yè)家論壇第6534期
主題：Inference on tree-structured subgroups with subgroup size and subgroup effect relationship in clinical trials在臨床試驗(yàn)中對(duì)具有子組大小和子組效應(yīng)關(guān)系的樹(shù)形結(jié)構(gòu)子組進(jìn)行推斷
主講人：香港科技大學(xué) 羅遠(yuǎn)暉 在讀博士生
主持人：統(tǒng)計(jì)學(xué)院 陳雪蓉教授
時(shí)間：5月24日 15:30-16:30
舉辦地點(diǎn)：柳林校區(qū)弘遠(yuǎn)樓408會(huì)議室
主辦單位：統(tǒng)計(jì)研究中心 統(tǒng)計(jì)學(xué)院 科研處
主講人簡(jiǎn)介：
羅遠(yuǎn)暉，香港科技大學(xué)在讀博士生。
內(nèi)容簡(jiǎn)介：
Subgroup analysis is frequently used to uncover and confirm treatment effect heterogeneity in clinical trials. When multiple candidate subgroups are considered, we often need to make statistical inference on the subgroups simultaneously. Classical multiple testing procedures might suffer from the loss of interpretability and efficiency as they often fail to take subgroup size and subgroup effect relationship into account. In this talk, built on the selective traversed accumulation rules (STAR), we propose a data-adaptive and interactive multiple testing procedure for subgroups which can take subgroup size and subgroup effect relationship into account under tree structure. The proposed method is easy-to-implement and can lead to a more efficient and interpretable inference on tree-structured subgroups. We demonstrate the merit of our proposed method by re-analyzing the panitumumab trial with the proposed method. This talk is based on joint work with Prof. Xinzhou Guo (HKUST).
子組分析經(jīng)常用于在臨床試驗(yàn)中揭示和確認(rèn)治療效應(yīng)的異質(zhì)性。當(dāng)考慮多個(gè)候選子組時(shí)，我們通常需要同時(shí)對(duì)子組進(jìn)行統(tǒng)計(jì)推斷。經(jīng)典的多重檢驗(yàn)程序可能會(huì)導(dǎo)致解釋性和效率的損失，因?yàn)樗鼈兺ǔＮ纯紤]子組大小和子組效應(yīng)關(guān)系。在本報(bào)告中，基于選擇性遍歷累積規(guī)則（STAR），我們提出了一種數(shù)據(jù)自適應(yīng)和交互式的多重檢驗(yàn)程序，用于考慮樹(shù)結(jié)構(gòu)下的子組大小和子組效應(yīng)關(guān)系。所提出的方法易于實(shí)現(xiàn)，并可以實(shí)現(xiàn)對(duì)樹(shù)形結(jié)構(gòu)子組的更高效和可解釋的推斷。我們通過(guò)使用所提出的方法重新分析泛替單抗試驗(yàn)來(lái)展示我們提出的方法的優(yōu)點(diǎn)。本報(bào)告基于與XinzhouGuo教授（香港科技大學(xué)）的合作研究。
