文献速递流行病学与生物统计

稀有事件荟萃分析中置信分布方法的比较

A comparison of confidence distribution approaches for rare event meta-analysis.
PubMed 文献 · Stat Med · 2021-07-04

针对稀有事件数据对传统荟萃分析方法的挑战,本文综述并比较了多种置信分布(CD)方法,包括Fisher经典P值合并法、P值函数合并法、置信区间合并法和置信对数似然函数合并法,并提出了这些方法的变体。研究发现,对于同质性稀有事件数据,大多数CD方法表现良好;对于异质性稀有事件数据,部分CD方法表现差异明显,一些表现很差而另一些尚可。

要点

  • 同质性稀有事件数据下,多数置信分布方法表现良好。
  • 异质性稀有事件数据下,不同置信分布方法性能差异明显。
  • 需根据数据同质性选择适宜的置信分布合并方法。

核心结论

  • 同质性稀有事件数据下多数CD方法表现良好;异质性数据下部分CD方法表现很差,另一些尚可。

研究设计与样本

  • 方法学比较研究,综述并比较Fisher经典P值合并法、P值函数合并法、置信区间合并法和置信对数似然函数合并法及其变体。

主要发现

  • 同质性稀有事件数据下,大多数CD方法表现很好。
  • 异质性稀有事件数据下,某些CD方法表现很差,另一些表现尚可。

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

Meta-analysis of rare event data has recently received increasing attention due to the challenging issues rare events pose to traditional meta-analytic methods. One specific way to combine information and analyze rare event meta-analysis data utilizes confidence distributions (CDs). While several CD methods exist, no comparisons have been made to determine which method is best suited for homogeneous or heterogeneous meta-analyses with rare events. In this article, we review several CD methods: Fisher's classic P-value combination method, one that combines P-value functions, another that combines confidence intervals, and one that combines confidence log-likelihood functions. We compare these CD approaches, and we propose and compare variations of these methods to determine which method produces reliable results for homogeneous or heterogeneous rare event meta-analyses. We find that for homogeneous rare event data, most CD methods perform very well. On the other hand, for heterogeneous rare event data, there is a clear split in performance between some CD methods, with some performing very poorly and others performing reasonably well.

来源:Stat Med,查看原文。版权归原作者所有。

引用

  1. Zabriskie BN, Corcoran C, Senchaudhuri P, et al. A comparison of confidence distribution approaches for rare event meta-analysis. Stat Med. 2021. PMID: 34219258. doi:10.1002/sim.9125

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。