Ciolino Jody D, Kaizer Alexander M, Bonner Lauren Balmert
Department of Preventive Medicine (Biostatistics), Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA.
Department of Biostatistics & Informatics, Colorado School of Public Health, Aurora, Colorado, USA.
J Clin Transl Sci. 2023 May 15;7(1):e124. doi: 10.1017/cts.2023.552. eCollection 2023.
Interim analyses in clinical trials can take on a multitude of forms. They are often used to guide Data and Safety Monitoring Board (DSMB) recommendations to study teams regarding recruitment targets for large, later-phase clinical trials. As collaborative biostatisticians working and teaching in multiple fields of research and across a broad array of trial phases, we note the large heterogeneity and confusion surrounding interim analyses in clinical trials. Thus, in this paper, we aim to provide a general overview and guidance on interim analyses for a nonstatistical audience. We explain each of the following types of interim analyses: efficacy, futility, safety, and sample size re-estimation, and we provide the reader with reasoning, examples, and implications for each. We emphasize that while the types of interim analyses employed may differ depending on the nature of the study, we would always recommend prespecification of the interim analytic plan to the extent possible with risk mitigation and trial integrity remaining a priority. Finally, we posit that interim analyses should be used as tools to help the DSMB make informed decisions in the context of the overarching study. They should generally not be deemed binding, and they should not be reviewed in isolation.
临床试验中的期中分析可以有多种形式。它们经常被用于指导数据与安全监测委员会（DSMB）就大型后期临床试验的招募目标向研究团队提出建议。作为在多个研究领域和广泛的试验阶段开展工作并进行教学的合作生物统计学家，我们注意到围绕临床试验期中分析存在很大的异质性和混乱。因此，在本文中，我们旨在为非统计学专业的读者提供关于期中分析的总体概述和指导。我们解释以下几种类型的期中分析：疗效、无效性、安全性和样本量重新估计，并为读者提供每种分析的理由、示例及影响。我们强调，虽然所采用的期中分析类型可能因研究性质而异，但我们始终建议尽可能预先制定期中分析计划，同时将风险缓解和试验完整性作为首要任务。最后，我们认为期中分析应作为工具，帮助DSMB在整体研究背景下做出明智决策。它们通常不应被视为具有约束力，也不应孤立地进行审查。
