Ge Lizhao, Hamasaki Toshimitsu, Evans Scott R
The Biostatistics Center, Milken Institute School of Public Health, The George Washington University, Rockville, MD, USA.
Department of Biostatistics and Bioinformatics, Milken Institute School of Public Health, The George Washington University, Washington, DC, USA.
Ther Innov Regul Sci. 2025 Mar;59(2):234-244. doi: 10.1007/s43441-024-00720-8. Epub 2024 Dec 17.
A data monitoring committee (DMC) can have an extremely challenging job. Stop a trial too soon, and results are inconclusive and the trial fails to obtain answers to important questions that could inform future clinical practice. Stop a trial too late, and trial participants are exposed to potentially harmful or ineffective interventions longer than necessary. Securing convincing and conclusive evidence and the ethical responsibility to current and future patients are weighed carefully during DMC deliberations. The ability to interpret complex information, and appreciation of issues affecting scientific integrity, are critical for the DMC to protect trial participants and public trust. Challenges faced by and issues of prudence faced by DMCs are discussed including interim analysis issues, assessing the totality of information with statistical boundaries as guidelines, interpretation of composite and surrogate outcomes, reactions to early trends, benefit:risk assessment, landscape changes, subgroup analyses, composing information for a comprehensive understanding of patient-centric effects, and evaluating the value of additional data. Case studies illustrate how DMCs addressed the challenges.
数据监测委员会（DMC）的工作极具挑战性。过早终止试验，结果将缺乏定论，试验也无法获得可为未来临床实践提供依据的重要问题的答案。过晚终止试验，试验参与者暴露于潜在有害或无效干预措施下的时间将超过必要时长。在DMC的审议过程中，需仔细权衡获取令人信服的确凿证据以及对当前和未来患者的伦理责任。解读复杂信息的能力以及对影响科学诚信问题的认识，对于DMC保护试验参与者和公众信任至关重要。本文将讨论DMC面临的挑战和审慎问题，包括期中分析问题、以统计边界为指导评估信息全貌、复合和替代结局的解读、对早期趋势的反应、获益-风险评估、情况变化、亚组分析、整合信息以全面了解以患者为中心的效应，以及评估额外数据的价值。案例研究说明了DMC如何应对这些挑战。
