文献速递流行病学与生物统计医学教育与科研方法

在改善痴呆症照护的实用临床试验中,采用贝叶斯多因素设计与分析估计多种干预措施的联合效应

A Bayesian Multi-Factorial Design and Analysis for Estimating Combined Effects of Multiple Interventions in a Pragmatic Clinical Trial to Improve Dementia Care.
PubMed 文献 · Stat Med · 2025-01-01

本文描述在大型整群随机试验ED-LEAD中,如何开发析因设计及贝叶斯分析方案,以估计多种干预措施对痴呆症患者照护改善的联合效应。文章讨论了析因设计中随机化单元可接受任意干预组合,以及研究问题对分析方式的影响。

要点

  • 析因设计可评估多种干预措施的联合效应。
  • 随机化单元可分配至干预措施的任意组合。
  • 研究问题决定析因设计及贝叶斯分析方案。

核心结论

  • 本文描述了在ED-LEAD大型整群随机试验中开发析因设计与贝叶斯分析方案的过程,用于估计多种干预措施的联合效应。

研究设计与样本

  • 研究设计为大型整群随机试验(ED-LEAD研究),采用析因设计与贝叶斯分析方案。

主要发现

  • 析因设计中随机化单元可被分配至干预措施的任意组合。
  • 采用析因设计及贝叶斯分析方案的理由因研究针对的具体问题而异,这些问题进而影响分析方式。

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

Factorial study designs can be important for understanding the effectiveness of interventions when multiple interventions are under investigation. In this design setting, a unit of randomization can be assigned to any combination of interventions. The rationale for taking this kind of approach can vary depending on the specific questions targeted by the research. These questions, in turn, have implications for the way in which the analyses will be conducted. The goal in this paper is to describe how we developed a factorial design along with a Bayesian analytic plan for a large cluster-randomized trial-the Emergency Departments LEading the transformation of Alzheimer's and Dementia care (ED-LEAD) study-focused on improving care for persons living with dementia.

来源:Stat Med,查看原文。版权归原作者所有。

引用

  1. Goldfeld KS, Grudzen CR, Shah MN, et al. A Bayesian Multi-Factorial Design and Analysis for Estimating Combined Effects of Multiple Interventions in a Pragmatic Clinical Trial to Improve Dementia Care. Stat Med. 2025. PMID: 40916513. doi:10.1002/sim.70264

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。