Prager Ross, Gagnon Luke, Bowdridge Joshua, Unni Rudy R, McGrath Trevor A, Cobey Kelly, Bossuyt Patrick M, McInnes Matthew D F
Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada
Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada.
BMJ Evid Based Med. 2021 Jan 22. doi: 10.1136/bmjebm-2020-111604.
Although the literature supporting the use of point-of-care ultrasound (POCUS) continues to grow, incomplete reporting of primary diagnostic accuracy studies has previously been identified as a barrier to translating research into practice and to performing unbiased systematic reviews. This study assesses POCUS investigator and journal editor attitudes towards barriers to adhering to the Standards for Reporting of Diagnostic Accuracy Studies (STARD) 2015 guidelines.
DESIGN, SETTING, PARTICIPANTS: Two separate surveys using a 5-point Likert scale were sent to POCUS study investigators and journal editors to assess for knowledge, attitude and behavioural barriers to the complete reporting of POCUS research. Respondents were identified based on a previous study assessing STARD 2015 adherence for POCUS studies published in emergency medicine, anaesthesia and critical care journals. Responses were anonymously linked to STARD 2015 adherence data from the previous study. Written responses were thematically grouped into the following categories: knowledge, attitude and behavioural barriers to quality reporting, or other. Likert response items are reported as median with IQRs.
The primary outcome was the median Likert score for the investigator and editor surveys assessing knowledge, attitude and behavioural beliefs about barriers to adhering to the STARD 2015 guidelines.
The investigator survey response rate was 18/69 (26%) and the editor response rate was 5/21 (24%). Most investigator respondents were emergency medicine practitioners (13/21, 62%). Two-thirds of investigators were aware of the STARD 2015 guidelines (12/18, 67%) and overall agreed that incomplete reporting limits generalisability and the ability to detect risk of bias (median 4 (4, 5)). Investigators felt that the STARD 2015 guidelines were useful, easy to find and easy to use (median 4 (4, 4.25); median 4 (4, 4.25) and median 4 (3, 4), respectively). There was a shared opinion held by investigators and editors that the peer review process be primarily responsible for ensuring complete research reporting (median 4 (3, 4) and median 4 (3.75, 4), respectively). Three of 18 authors (17%) felt that the English publication language of STARD 2015 was a barrier to adherence.
Although investigators and editors recognise the importance of completely reported research, reporting quality is still a core issue for POCUS research. The shared opinion held by investigators and editors that the peer review process be primarily responsible for reporting quality is potentially problematic; we view completely reported research as an integral part of the research process that investigators are responsible for, with the peer review process serving as another additional layer of quality control. Endorsement of reporting guidelines by journals, auditing reporting guideline adherence during the peer review process and translation of STARD 2015 guidelines into additional languages may improve reporting completeness for the acute POCUS literature.
Open Science Framework Registry (https://osf.io/5pzxs/).
尽管支持使用床旁超声（POCUS）的文献不断增加，但先前已发现初级诊断准确性研究的报告不完整是将研究转化为实践以及进行无偏倚系统评价的障碍。本研究评估了POCUS研究者和期刊编辑对遵守《2015年诊断准确性研究报告标准》（STARD）指南的障碍的态度。
设计、背景、参与者：使用5点李克特量表分别向POCUS研究的研究者和期刊编辑发送了两份单独的调查问卷，以评估POCUS研究完整报告的知识、态度和行为障碍。根据之前一项评估急诊医学、麻醉学和重症监护期刊上发表的POCUS研究对STARD 2015的遵守情况的研究来确定受访者。回复与之前研究中的STARD 2015遵守数据进行匿名关联。书面回复按主题分为以下几类：质量报告的知识、态度和行为障碍，或其他。李克特回答项目报告为中位数及四分位间距。
主要结局是研究者和编辑调查问卷的中位数李克特评分，评估对遵守STARD 2015指南障碍的知识、态度和行为信念。
研究者调查问卷的回复率为18/69（26%），编辑的回复率为5/21（24%）。大多数研究者受访者是急诊医学从业者（13/21，62%）。三分之二的研究者知晓STARD 2015指南（12/18，67%），总体上同意报告不完整会限制普遍性以及检测偏倚风险的能力（中位数4（4，5））。研究者认为STARD 2015指南有用、易于查找且易于使用（中位数分别为4（4，4.25）；4（4，4.25）和4（3，4））。研究者和编辑有一个共同的观点，即同行评审过程应主要负责确保研究报告完整（中位数分别为4（3，4）和4（3.75，4））。18位作者中有3位（17%）认为STARD 2015的英文出版语言是遵守的障碍。
尽管研究者和编辑认识到完整报告研究的重要性，但报告质量仍是POCUS研究的核心问题。研究者和编辑共同认为同行评审过程应主要负责报告质量这一观点可能存在问题；我们认为完整报告的研究是研究过程中研究者负责的一个组成部分，同行评审过程是另一层额外的质量控制。期刊对报告指南的认可、在同行评审过程中审核报告指南的遵守情况以及将STARD 2015指南翻译成其他语言可能会提高急性POCUS文献的报告完整性。
开放科学框架注册库（https://osf.io/5pzxs/）。
