Sumner Petroc, Schwartz Lisa, Woloshin Steven, Bratton Luke, Chambers Christopher
School of Psychology, Cardiff University, Cardiff, UK
Center for Medicine and the Media, The Dartmouth Institute for Health Policy and Clinical Practice, Lebanon, New Hampshire, USA.
BMJ Open. 2021 Jan 8;11(1):e041385. doi: 10.1136/bmjopen-2020-041385.
To examine how often study funding and author conflicts of interest are stated in science and health press releases and in corresponding news; and whether disclosure in press releases is associated with disclosure in news. Second, to specifically examine disclosure rates in industry-funded studies.
Retrospective content analysis with two cohorts.
Press releases about health, psychology or neuroscience research from research universities and journals from 2011 (n=996) and 2015 (n=254) and their associated news stories (n=1250 and 578).
Mention of study funding and author conflicts of interest.
In our 2011 cohort, funding was reported in 94% (934/996) of journal articles, 29% (284/996) of press releases and 9% (112/1250) of news. The corresponding figures for 2015 were: 84% (214/254), 52% (131/254) and 10% (58/578). A similar pattern was seen for the industry funding subset. If the press release reported study funding, news was more likely to: 22% if in the press release versus 7% if not in the press release (2011), relative risk (RR) 3.1 (95% CI 2.1 to 4.3); for 2015, corresponding figures were 16% versus 2%, RR 6.8 (95% CI 2.2 to 17). In journal articles, 27% and 22% reported a conflict of interest, while less than 2% of press releases or news ever mentioned these.
Press releases and associated news did not frequently state funding sources or conflicts of interest. Funding information in press releases was associated with such information in news. Given converging evidence that news draws on press release content, including statements of funding and conflicts of interest in press releases may lead to increased reporting in news.
研究科学与健康新闻稿及相应新闻中提及研究资金和作者利益冲突的频率；以及新闻稿中的披露情况是否与新闻中的披露相关。其次，具体研究行业资助研究中的披露率。
对两个队列进行回顾性内容分析。
来自研究型大学和期刊的关于健康、心理学或神经科学研究的新闻稿，2011年有996篇，2015年有254篇，以及它们相关的新闻报道（分别为1250篇和578篇）。
提及研究资金和作者利益冲突。
在我们2011年的队列中，94%（934/996）的期刊文章、29%（284/996）的新闻稿和9%（112/1250）的新闻报道了资金情况。2015年的相应数据为：84%（214/254）、52%（131/254）和10%（58/578）。行业资助子集也呈现类似模式。如果新闻稿报道了研究资金，新闻报道的可能性更高：2011年，新闻稿中报道的为22%，未报道的为7%，相对风险（RR）为3.1（95%置信区间2.1至4.3）；2015年，相应数据为16%对2%，RR为6.8（95%置信区间2.2至17）。在期刊文章中，27%和22%报告了利益冲突，而新闻稿或新闻中提及这些的不到2%。
新闻稿及相关新闻并未频繁说明资金来源或利益冲突。新闻稿中的资金信息与新闻中的此类信息相关。鉴于越来越多的证据表明新闻借鉴了新闻稿内容，在新闻稿中纳入资金和利益冲突声明可能会增加新闻报道。
