Korfitsen Christoffer Bruun, Nejstgaard Camilla Hansen, Hróbjartsson Asbjørn, Boutron Isabelle, Bero Lisa, Lundh Andreas
Cochrane Denmark & Centre for Evidence-Based Medicine Odense (CEBMO), Department of Clinical Research, University of Southern Denmark, Odense, Denmark
Open Patient Data Explorative Network, Odense University Hospital, Odense, Syddanmark, Denmark.
BMJ Evid Based Med. 2025 Mar 21;30(2):104-117. doi: 10.1136/bmjebm-2024-112967.
Peer review may improve the quality of research manuscripts and aid in editorial decisions, but reviewers can have conflicts of interest that impact on their recommendations.
The objective was to systematically map and describe the extent and nature of empirical research on peer reviewers' conflicts of interest in biomedical research.
Scoping review METHODS: In this scoping review, we included studies investigating peer reviewers' conflicts of interest in journal manuscripts, theses and dissertations, conference abstracts, funding applications and clinical guidelines. We searched MEDLINE, Embase, The Cochrane Methodology Register, Google Scholar (up to January 2024) and other sources. Two authors independently included studies and extracted data on key study characteristics and results, and we organised data by study domain (eg, journal manuscripts) and study aims. We included studies directly investigating peer reviewers' conflicts of interest in our primary analysis, and studies investigating other questions (eg, reasons for retraction), but reporting relevant data on peer reviewers' conflicts of interest, were solely included in a supplementary analysis.
After screening 44 353 references, we included 71 studies, of which 41 were included in our primary analysis. The 41 studies were published between 2005 and 2023, and 34 (83%) were journal publications. 30 (73%) studies investigated journal manuscripts, 1 (2%) conference abstracts, 4 (10%) funding applications and 6 (15%) clinical guidelines. No studies investigated theses or dissertations. 37 (90%) studies used quantitative research methods, 2 (5%) qualitative and 2 (5%) mixed methods. 21 (51%) studies investigated both financial and non-financial interests, 6 (15%) solely financial interests, 5 (12%) solely non-financial interests and 9 (22%) did not report the type of interest. We organised included studies based on study aims, with some studies having multiple aims: impact on recommendations (one study), occurrence of peer reviewers' conflicts of interest (11 studies), stakeholders' experiences (13 studies) and policy and management (22 studies). One (2%) study investigated the impact of peer reviewers' personal connections with authors on reviewers' recommendations. Nine (22%) studies estimated prevalences of conflicts of interest among peer reviewers, ranging from 3%-91%. Two (5%) studies both reported that conflicts of interest were a reason for declining to review in 1% of cases. 13 (32%) studies investigated stakeholders' experiences with peer reviewers' conflicts of interest, primarily using questionnaires of reviewers, editors and researchers. 16 (39%) studies estimated prevalences of having conflict of interest policies for peer reviewers, ranging from 5%-96%, among journals, conferences and clinical guideline organisations. Finally, six (15%) studies estimated prevalences of public availabilities of reviewers' conflicts of interest declarations, ranging from 0%-71%.
Most studies addressed conflicts of interest in peer review of journal manuscripts, primarily through surveys of journal policies or questionnaires of researchers, editors and peer reviewers. The impact of peer reviewers' conflicts of interest on recommendations and their prevalence is still poorly understood. Our results can guide future studies and be used to align policies and management of peer reviewers' conflicts of interest.
Open Science Framework (DOI: 10.17605/OSF.IO/9QBMG).
同行评审可能会提高研究手稿的质量，并有助于编辑决策，但评审人员可能存在利益冲突，从而影响他们的建议。
本研究旨在系统梳理和描述关于生物医学研究中同行评审人员利益冲突的实证研究范围及性质。
范围综述
在本范围综述中，我们纳入了调查同行评审人员在期刊手稿、学位论文、会议摘要、基金申请和临床指南中的利益冲突的研究。我们检索了MEDLINE、Embase、Cochrane方法学注册库、谷歌学术（截至2024年1月）及其他来源。两位作者独立纳入研究，并提取关于关键研究特征和结果的数据，我们按研究领域（如期刊手稿）和研究目的对数据进行整理。我们将直接调查同行评审人员利益冲突的研究纳入主要分析，将调查其他问题（如撤稿原因）但报告了同行评审人员利益冲突相关数据的研究仅纳入补充分析。
在筛选了44353篇参考文献后我们纳入了71项研究，其中41项纳入主要分析。这41项研究发表于2005年至2023年之间，34项（83%）为期刊出版物。30项（73%）研究调查期刊手稿，1项（2%）会议摘要，4项（10%）基金申请，6项（15%）临床指南。没有研究调查学位论文。37项（90%）研究采用定量研究方法，2项（5%）定性研究方法，2项（5%）混合研究方法。21项（51%）研究调查了经济利益和非经济利益，6项（15%）仅调查经济利益，5项（12%）仅调查非经济利益，9项（22%）未报告利益类型。我们根据研究目的对纳入研究进行整理，有些研究有多个目的：对建议的影响（1项研究）、同行评审人员利益冲突的发生情况（11项研究）、利益相关者的经历（13项研究）以及政策与管理（22项研究）。1项（2%）研究调查了同行评审人员与作者的个人关系对评审人员建议的影响。9项（22%）研究估计了同行评审人员中利益冲突的发生率，范围为3% - 91%。2项（5%）研究均报告称利益冲突在1%的情况下是拒绝评审的原因。13项（32%）研究调查了利益相关者对同行评审人员利益冲突的经历，主要通过对评审人员、编辑和研究人员进行问卷调查。16项（39%）研究估计了期刊、会议和临床指南组织中针对同行评审人员的利益冲突政策的发生率，范围为5% - 96%。最后，6项（15%）研究估计了评审人员利益冲突声明的公开可得率，范围为0% - 71%。
大多数研究关注期刊手稿同行评审中的利益冲突，主要通过调查期刊政策或对研究人员、编辑和同行评审人员进行问卷调查。同行评审人员利益冲突对建议的影响及其发生率仍了解不足。我们的结果可为未来研究提供指导，并用于协调同行评审人员利益冲突的政策与管理。
开放科学框架（DOI：10.17605/OSF.IO/9QBMG）
