Are current standards of reporting quality for clinical trials sufficient in addressing important sources of bias?
Are current standards of reporting quality for clinical trials sufficient in addressing important sources of bias?
复制标题
DOI：
10.1016/j.cct.2015.07.019
复制
发表时间：
2015-11-01
影响因子：
2.2
通讯作者：
Thorlund, Kristian
中科院分区：
文献类型：
--
作者：
Mills, Edward J.;Ayers, Dieter;Thorlund, Kristian
确定随机临床试验 (RCT) 的质量对于决策者确定试验结果的可信度和适用性来说是必要的。可能影响 RCT 证据效用的问题包括偏倚、随机误差和适用性问题。在本文中，我们主要关注偏见问题，并检查报告方法项目（包括分配隐藏、序列生成和参与者盲法）是否可以作为偏见证据的证据。我们提出了对元流行病学研究的系统回顾和一项模拟研究的结果，表明经常检查的偏差来源可能在治疗夸大中发挥很小的作用。我们讨论可能额外影响试验结果的其他问题，包括样本量、发表偏倚和试验人员的专业知识。最后，我们讨论了在评估治疗效果总体估计时调节已知偏差影响的策略。 (C) 2015 Elsevier Inc. 保留所有权利。
Determining the quality of a randomized clinical trial (RCT) is necessary for decision-makers to determine the believability and applicability of the trial findings. Issues that are likely to affect the utility of RCT evidence include issues of bias, random error and applicability. In this article we focus primarily on issues of bias and examine the evidence for whether reporting methodological items, including allocation concealment, sequence generation, and blinding of participants can be relied upon as evidence of bias. We present the findings of a systematic review of meta-epidemiological studies and a simulation study demonstrating that commonly examined sources of bias likely play little role in treatment exaggeration. We discuss other issues that may additionally influence trial outcomes including sample size, publication bias, and expertise of trialists. We conclude by discussing strategies to moderate the effect of known biases in assessing overall estimates of treatment effects. (C) 2015 Elsevier Inc. All rights reserved.