平台试验的特征、设计与统计方法:一项系统综述
该系统综述从试验注册库和文献数据库检索至2024年8月,共纳入189项平台试验,主要涉及感染性疾病(77项,其中57项为COVID-19)和肿瘤(68项)。研究发现平台试验启动在COVID-19大流行期间达到高峰后趋于稳定,47%的试验有完整主方案,贝叶斯设计58项、频率学派设计56项,23%的试验未预设目标样本量且均为贝叶斯设计,模拟报告透明度有限,仅2项试验共享模拟代码。
要点
- 共纳入189项平台试验,感染性疾病77项(含COVID-19 57项)和肿瘤68项为主要领域。
- 47%(89/189)的平台试验可获得完整主方案;贝叶斯设计58/189项,频率学派56/189项,两者兼用20/189项。
- 25/111项(23%)试验未预设目标样本量且均为贝叶斯设计,其中16项明确报告为「永久」试验;仅2项试验共享模拟代码。
核心结论
- 平台试验仍受欢迎且日益多样化,需提升透明度和报告质量,尤其在复杂贝叶斯平台试验中,以确保可靠性和更广泛接受度。
研究设计与样本
- 系统综述,从试验注册库和文献数据库检索截至2024年8月的平台试验。
- 纳入标准为在单一人群中研究多种干预、可灵活增加或删除试验组的随机对照试验,共纳入189项平台试验。
- 提取数据包括试验状态、设计、统计方法和报告实践。
主要发现
- 189项平台试验中,感染性疾病77项(其中COVID-19 57项),肿瘤68项。
- 平台试验启动在COVID-19大流行期间达到高峰,此后稳定在84项活跃试验,另有25项在计划中。
- 47%(89/189)的平台试验可获得完整主方案;贝叶斯设计58/189项 vs 频率学派56/189项,两者兼用20/189项(55/189项不明确)。
- 25/111项(23%)试验设计未预设目标样本量,均为贝叶斯设计,其中16项明确报告为「永久」试验。
- 中期分析次数预设:贝叶斯试验18%(10/57) vs 频率学派试验58%(28/48);贝叶斯试验中93%(39/42)使用模拟评估操作特征,67%(26/39)可获得模拟报告,62%(24/39)详细描述了模拟程序,仅2项试验共享模拟代码。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND AND OBJECTIVE: Platform trials (PTs) are gaining popularity in clinical research due to their innovative and flexible methodologies. The objective was to determine the characteristics, methodological, and statistical practices in PTs. METHODS: We identified PTs from trial registries and bibliographic databases up to August 2024. Eligible PTs were randomized controlled trials studying multiple interventions within a single population, with flexibility to add or drop arms. Data were extracted on trial status, design, statistical methods, and reporting practices. RESULTS: We identified 189 PTs. Most focused on infectious diseases (77, including 57 for COVID-19) and oncology (68). PT initiation peaked during the COVID-19 pandemic but has since stabilized at 84 active PTs, with 25 in planning. A complete master protocol was available for 47% (89/189) of PTs. Bayesian designs featured in 58/189 PTs vs. 56/189 frequentist trials, 20/189 trials utilizing both (unclear in 55/189 PTs). Overall, 25/111 trials (23%) were designed without a predetermined target sample size, all of which were Bayesian. Among these, 16 were explicitly reported as "perpetual" trials. The number of interim analyses was predetermined in 18% (10/57) of Bayesian trials vs. 58% (28/48) of frequentist trials. Simulations to evaluate operating characteristics were used in 93% (39/42) of Bayesian trials. Simulation reports were available in 67% (26/39) of cases, and the procedures were detailed for 62% (24/39) of trials. Only two trials shared the simulation code. CONCLUSION: PTs remain popular and increasingly diverse. Efforts to enhance transparency and reporting, especially in complex Bayesian PTs, are essential to ensure reliability and broader acceptance.
引用
- Massonnaud CR, Schönenberger CM, Chiaborelli M, et al. Characteristics, design, and statistical methods in platform trials: a systematic review. J Clin Epidemiol. 2025. PMID: 40349734. doi:10.1016/j.jclinepi.2025.111827
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。