Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China.
Department of Public Health and Primary Care, School of Clinical Medicine, University of Cambridge, Cambridge, United Kingdom.
JAMA Netw Open. 2023 Mar 1;6(3):e234219. doi: 10.1001/jamanetworkopen.2023.4219.
Systematic reviews can help to justify a new randomized clinical trial (RCT), inform its design, and interpret its results in the context of prior evidence.
To assess trends and factors associated with citing (a marker of the use of) prior systematic reviews in RCT reports.
DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study investigated 737 Cochrane reviews assessing health interventions to identify 4003 eligible RCTs, defined as those included in an updated version but not in the first version of a Cochrane review and published 2 years after the first version of the Cochrane review was published.
The primary outcome was the citation of prior systematic reviews, Cochrane or others, as determined by screening references of eligible RCTs. Factors that may be associated with the citation of prior systematic reviews were also examined.
Among 4003 eligible RCTs, 1241 studies (31.0%) cited Cochrane reviews, 1698 studies (42.4%) cited prior non-Cochrane reviews, and 2265 studies (56.6%) cited either type of systematic review or both; 1738 RCTs (43.4%) cited no systematic reviews. The percentage of RCTs citing prior Cochrane reviews, non-Cochrane reviews, and either or both types of review increased from 28 studies (15.3%), 46 studies (25.1%), and 65 studies (35.5%) of 183 RCTs before 2008 to 42 studies (40.8%), 65 studies (64.1%), and 73 studies (71.8%) of 102 RCTs since 2020, respectively; the annual increases were 1.9% (95% CI, 1.4%-2.3%), 3.3% (95% CI, 2.9%-3.7%), and 3.0% (95% CI, 2.5%-3.5%), respectively. The proportion of RCTs citating prior systematic reviews varied considerably across clinical specialties, ranging from 28 of 106 RCTs (26.4%) in ophthalmology to 386 of 553 RCTs (69.8%) in psychiatry (P < .001). RCTs with 100 participants or more (risk ratio [RR], 1.16; 95% CI, 1.03-1.30), nonindustry funding (RR, 1.43; 95% CI, 1.27-1.61), and authors from high-income countries (RR, 1.10; 95% CI, 1.03-1.17) were more likely to cite systematic reviews than those with fewer than 100 participants, industry funding, and authors from low- and middle-income countries, respectively. A journal requirement to cite systematic reviews was not associated with the likelihood of citing a systematic review.
This study found that the citation of prior systematic reviews in RCT reports improved over time, but approximately 40% of RCTs failed to do so. These findings suggest that reference to prior evidence for initiating, designing, and reporting RCTs should be further emphasized to assure clinical relevance, improve methodological quality, and facilitate interpretation of new results.
系统评价有助于为新的随机临床试验（RCT）提供依据，为其设计提供信息，并在先前证据的背景下解释其结果。
评估引用（使用的标志）RCT 报告中先前系统评价的趋势和相关因素。
设计、设置和参与者：本横断面研究调查了 737 项评估健康干预措施的 Cochrane 评价，以确定 4003 项合格的 RCT，定义为那些包含在更新版本中但不包含在 Cochrane 评价的第一版本中的 RCT，并且在 Cochrane 评价的第一版本发布后 2 年发布。
主要结果是通过筛选合格 RCT 的参考文献来确定先前系统评价的引用情况，包括 Cochrane 或其他系统评价。还检查了可能与引用先前系统评价相关的因素。
在 4003 项合格的 RCT 中，1241 项研究（31.0%）引用了 Cochrane 评价，1698 项研究（42.4%）引用了先前的非 Cochrane 评价，2265 项研究（56.6%）引用了这两种类型的系统评价或两者；1738 项 RCT（43.4%）未引用系统评价。引用先前 Cochrane 评价、非 Cochrane 评价和这两种类型的评价的 RCT 比例从 2008 年前的 183 项 RCT 中的 28 项（15.3%）、46 项（25.1%）和 65 项（35.5%）分别增加到 2020 年后的 102 项 RCT 中的 42 项（40.8%）、65 项（64.1%）和 73 项（71.8%）；年增长率分别为 1.9%（95%CI，1.4%-2.3%）、3.3%（95%CI，2.9%-3.7%）和 3.0%（95%CI，2.5%-3.5%）。引用先前系统评价的 RCT 在临床专科之间差异很大，从眼科的 106 项 RCT 中的 28 项（26.4%）到精神病学的 553 项 RCT 中的 386 项（69.8%）（P < .001）。有 100 名或以上参与者（风险比[RR]，1.16；95%CI，1.03-1.30）、非行业资助（RR，1.43；95%CI，1.27-1.61）和来自高收入国家的作者（RR，1.10；95%CI，1.03-1.17）的 RCT 比参与者少于 100 人、有行业资助和来自中低收入国家的作者更有可能引用系统评价。要求引用系统评价的期刊与引用系统评价的可能性之间没有关联。
本研究发现，RCT 报告中引用先前系统评价的情况随着时间的推移有所改善，但仍有约 40%的 RCT 没有这样做。这些发现表明，为了确保临床相关性、提高方法学质量和促进对新结果的解释，应进一步强调参考先前的证据来启动、设计和报告 RCT。
