文献速递流行病学与生物统计

多重次要结局分析:精准解读至关重要

Multiple secondary outcome analyses: precise interpretation is important.
PubMed 文献 · Trials · 2022-01-10

本文讨论临床试验中多重次要结局分析的问题,质疑未做多重性校正即必须视为探索性的观点,认为只要对单个次要结局结果进行仔细精准解读,无需降级为探索性,因为每次比较的假阳性率并不随多重检验而增加,次要结局的强效应应始终被认真对待。

要点

  • 多重次要结局分析增加至少一项假阳性结果概率,但每次比较的假阳性率不随多重检验增加。
  • 未做多重性校正不必然使次要结局结果降级为探索性,精准解读单个结果即可。
  • 次要结局的强效应应始终被认真对待,不应仅因多重性担忧而否定。

核心结论

  • 若对单个次要结局结果进行仔细精准解读,无需将未做多重性校正的次要结局分析降级为探索性,次要结局的强效应应始终被认真对待。

研究设计与样本

  • 原文未报告研究设计。

主要发现

  • 多重次要结局分析增加至少一项假显著结果的概率,但每次比较的假显著结果概率(即每次比较的I类错误率)不随多重检验而增加。

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

Analysis of multiple secondary outcomes in a clinical trial leads to an increased probability of at least one false significant result among all secondary outcomes studied. In this paper, we question the notion that that if no multiplicity adjustment has been applied to multiple secondary outcome analyses in a clinical trial, then they must necessarily be regarded as exploratory. Instead, we argue that if individual secondary outcome results are interpreted carefully and precisely, there is no need to downgrade our interpretation to exploratory. This is because the probability of a false significant result for each comparison, the per-comparison wise error rate, does not increase with multiple testing. Strong effects on secondary outcomes should always be taken seriously and must not be dismissed purely on the basis of multiplicity concerns.

来源:Trials,查看原文。版权归原作者所有。

引用

  1. Parker RA, Weir CJ. Multiple secondary outcome analyses: precise interpretation is important. Trials. 2022. PMID: 35012627. doi:10.1186/s13063-021-05975-2

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。