Evaluation of spin in abstracts of papers in psychiatry and psychology journals
Evaluation of spin in abstracts of papers in psychiatry and psychology journals
复制标题
DOI：
10.1136/bmjebm-2019-111176
复制
发表时间：
2020-10-01
影响因子：
5.8
通讯作者：
Vassar, Matt
中科院分区：
文献类型：
--
作者：
Jellison, Samuel;Roberts, Will;Vassar, Matt
我们在精神病学和心理学期刊的随机对照试验（RCT）摘要中发现了“旋转”，其主要终点不显着。这是对 2012 年 1 月至 2017 年 12 月发表在精神病学和心理学期刊上的主要终点不显着的临床试验的横断面综述。主要结果是摘要中旋转的频率和表现。我们将旋转定义为“无论出于何种动机，使用特定的报告策略来强调实验治疗是有益的，尽管主要结果在统计上不显着，或者分散读者对统计上不显着的结果的注意力”。我们还评估了行业资金和旋转之间的关系。在检查的 486 项随机对照试验中，我们对旋转进行了分析，其中包括 116 项。其中 56% (n=65) 被识别为旋转。在 2 个 (2%) 个标题、24 个 (21%) 个摘要结果部分和 57 个 (49.1%) 个摘要结论部分中发现了旋转。在 15% 的随机对照试验 (n=17) 中，结果和结论部分同时发现了旋转的证据。 12 篇文章报道了行业资助（10%）。抽象而言，行业资助与旋转几率的增加无关（未调整 OR：1.0；95% CI：0.3 至 3.2）。我们在摘要中发现行业资助和旋转之间没有关系。这些发现引起了人们对旋转可能对临床医生产生的影响的担忧。可以采取进一步措施来解决虚假陈述问题，包括邀请审稿人对虚假陈述的存在发表评论，并更新《报告试验综合标准》指南以包含不鼓励虚假陈述的语言。
We have identified 'spin' in abstracts of randomised controlled trials (RCTs) with nonsignificant primary endpoints in psychiatry and psychology journals. This is a cross-sectional review of clinical trials with nonsignificant primary endpoints published in psychiatry and psychology journals from January 2012 to December 2017. The main outcome was the frequency and manifestation of spin in the abstracts. We define spin as the 'use of specific reporting strategies, from whatever motive, to highlight that the experimental treatment is beneficial, despite a statistically nonsignificant difference for the primary outcome, or to distract the reader from statistically nonsignificant results'. We have also assessed the relationship between industry funding and spin. Of the 486 RCTs examined, 116 were included in our analysis of spin. Spin was identified in 56% (n=65) of those included. Spin was found in 2 (2%) titles, 24 (21%) abstract results sections and 57 (49.1%) abstract conclusion sections. Evidence of spin was simultaneously identified in both results and conclusions sections in 15% of RCTs (n=17). Twelve articles reported industry funding (10%). Industry funding was not associated with increased odds of spin in the abstract (unadjusted OR: 1.0; 95% CI: 0.3 to 3.2). We found no relationship between industry funding and spin in abstracts. These findings raise concerns about the effects spin may have on clinicians. Further steps could be taken to address spin, including inviting reviewers to comment on the presence of spin and updating Consolidated Standards of Reporting Trials guidelines to contain language discouraging spin.