文献速递流行病学与生物统计临床试验与 GCP

小样本序贯多阶段随机试验(snSMART)的贝叶斯分析

A Bayesian analysis of small n sequential multiple assignment randomized trials (snSMARTs).
PubMed 文献 · Stat Med · 2018-07-16

针对罕见病临床试验患者有限的问题,snSMART设计将应答者继续原治疗、无应答者二次随机至剩余治疗,并利用两阶段数据比较疗效。作者提出一种贝叶斯方法实现两阶段信息借用,通过模拟比较偏差、均方根误差、区间宽度及95%置信/可信区间覆盖率,发现其估计量的均方根误差和区间宽度在现实设定下小于仅用第一阶段数据的方法及采用广义估计方程的对数泊松模型,表明利用第二阶段数据可改善罕见病治疗推断。

要点

  • snSMART设计适用于罕见病,通过两阶段随机和再随机利用全部患者数据。
  • 所提贝叶斯方法可借用两阶段信息,模拟中均方根误差和95%区间宽度更小。
  • 仅用第一阶段数据或对数泊松GEE模型在现实设定下推断效果不及所提方法。

核心结论

  • 在snSMART中收集并使用第二阶段数据可为罕见病治疗提供改进的推断。
  • 所提贝叶斯估计量的均方根误差和95%置信/可信区间宽度在现实设定下小于仅用第一阶段数据的方法及对数泊松GEE模型。

研究设计与样本

  • 研究方法为贝叶斯分析,通过模拟比较所提方法与传统方法的偏差、均方根误差、区间宽度和95%置信/可信区间覆盖率。
  • snSMART设计为:患者先随机分配至多种治疗之一(第一阶段),应答者继续原治疗,无应答者再随机分配至剩余治疗之一(第二阶段),两阶段数据均用于比较疗效。

主要发现

  • 所提贝叶斯估计量的均方根误差和95%置信/可信区间宽度在现实设定下小于仅用第一阶段数据的方法及对数泊松GEE模型。
  • 模拟比较了偏差、均方根误差、区间宽度和95%置信/可信区间覆盖率等指标。

临床怎么用

  • 在罕见病临床试验中,可考虑采用snSMART设计并利用贝叶斯方法借用两阶段信息以提高推断效率。

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

Designing clinical trials to study treatments for rare diseases is challenging because of the limited number of available patients. A suggested design is known as the small n sequential multiple assignment randomized trial (snSMART), in which patients are first randomized to one of multiple treatments (stage 1). Patients who respond to their initial treatment continue the same treatment for another stage, while those who fail to respond are rerandomized to one of the remaining treatments (stage 2). The data from both stages are used to compare the efficacy between treatments. Analysis approaches for snSMARTs are limited, and we propose a Bayesian approach that allows for borrowing of information across both stages. Through simulation, we compare the bias, root-mean-square error, width, and coverage rate of 95% confidence/credible interval of estimators from of our approach to estimators produced from (i) standard approaches that only use the data from stage 1, and (ii) a log-Poisson model using data from both stages whose parameters are estimated via generalized estimating equations. We demonstrate the root-mean-square error and width of 95% confidence/credible intervals of our estimators are smaller than the other approaches in realistic settings, so that the collection and use of stage 2 data in snSMARTs provide improved inference for treatments of rare diseases.

来源:Stat Med,查看原文。版权归原作者所有。

引用

  1. Wei B, Braun TM, Tamura RN, et al. A Bayesian analysis of small n sequential multiple assignment randomized trials (snSMARTs). Stat Med. 2018. PMID: 30010207. doi:10.1002/sim.7900

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。