GoodReports: An online tool helping authors find and use reporting checklists to improve completeness of reporting

研究人员通过在期刊上撰写和发表报告来交流他们的工作和发现。然而，有人认为，大约80%的医学研究实际上并没有为我们的知识做出贡献，因此不能用于改善医疗保健。不适当的研究设计，延迟发表，不发表，在出版物中没有提供足够的细节，只发表一些结果是造成这种浪费的一些原因。对解决这一废物问题感兴趣的人们多年来一直在调查原因和可能的解决方案。资助者、期刊、研究人员和其他人都支持倡议，并开发了工具来帮助研究人员设计、运行和撰写他们的研究。报告指南就是这样一个工具。报告指南就像一个购物清单，提醒人们写文章的关键信息，读者需要重复研究或判断方法是否足够好，使结果可靠。最好的报告指南是由专家小组按照严格的程序制定的，并在同行评议的期刊上发表。报告指南在不同的时间在不同的期刊上发表和更新，这使得研究人员很难跟上可用的内容。这些工具的认知度很低，特别是在早期职业研究人员中，因为研究人员在培训期间没有经常被教导这些工具。好的期刊会推荐报告准则，有些期刊会强制要求作者提交一份填好的清单和他们的手稿。然而，通常无论是处理来稿的期刊编辑还是作者都不知道哪种报告准则是最合适的。作者通常需要不止一个指导方针。赤道网络致力于推广写得好的、完整的卫生研究文章，例如通过我们免费提供的、可全面搜索的报告指南数据库。我们开发了一个名为GoodReports的原型在线工具，以帮助研究人员更容易地找到和使用报告指南（www.goodreports.org）。为了表彰该工具在初步研究中的前景，它于2018年9月因减少研究浪费而获得了科克伦奖励奖。GoodReports在线工具向研究人员询问了一组关于他们研究的简单问题，以“诊断”他们需要哪种报告指南。该工具提供了推荐清单的在线版本，每个项目都有弹出式的解释，以及可用的良好报告示例。清单也可以下载以离线使用。GoodReports要求作者浏览清单，澄清或添加更多的信息到他们的手稿中，如果需要的话，标记他们在手稿中报告的信息，并提交填好的清单和他们的手稿。GoodReports原型包含16个报告准则。在四份BMC期刊上进行的试点发现，使用GoodReports的作者更有可能使用正确的报告指南。一项针对BMJ Open作者的研究的初步数据表明，使用该工具的作者比不使用该工具的作者更有可能在他们的手稿中提交一份报告清单。我们建议将GoodReports从原型改进为可共享的工具。我们将添加更多的检查表，使用户能够组合来自多个报告指南的元素，并为每个检查表项添加至少一个良好报告的示例。开发将基于我们两项初步研究的数据和用户反馈。在整个项目中，我们将继续在线收集GoodReports用户反馈，以迭代地指导工具的开发。我们还将在更广泛的在线调查中询问当地焦点小组中早期职业研究人员的意见，以及作者和其他潜在用户（例如，期刊编辑和医学作者）的意见。然后，我们将在随机试验中测试GoodReports工具是否增加了提交给医学期刊的手稿的完整性。

Researchers communicate about their work and findings by writing and publishing reports in journals. However, it has been suggested that around 80% of all medical research doesn't actually contribute to our knowledge, so can't be used to improve healthcare. Inappropriate study design, delays in publishing, non-publishing, not giving enough detail in a publication, and only publishing some results are some reasons for this waste.People interested in solving this waste problem have been investigating causes and possible solutions for many years. Funders, journals, researchers, and others have supported initiatives and developed tools to help researchers design, run, and write up their research. Reporting guidelines are one such tool. A reporting guideline is like a shopping list to remind people writing articles of the key information that readers need to repeat the research or judge if the methods were good enough to make the results reliable. The best reporting guidelines are developed by expert groups following a rigorous procedure and written up and published in a peer-reviewed journal. Reporting guidelines are published and updated at different times in different journals, making it difficult for researchers to keep up to date with what's available. Awareness of these tools, particularly among early-career researchers, is low as researchers are not routinely taught about them during training. Good journals recommend reporting guidelines, and some make it compulsory for authors to submit a filled-in checklist with their manuscript. However, often neither the journal editors dealing with incoming manuscripts nor the authors know which reporting guideline is most appropriate. And often authors need more than one guideline.The EQUATOR Network works to promote well-written, complete health research articles, for example through our freely available, comprehensive searchable database of reporting guidelines. We developed a prototype online tool called GoodReports to help researchers find and use reporting guidelines more easily (www.goodreports.org). In recognition of the tool's promise in preliminary studies, it won the Cochrane-REWARD prize for reducing research waste in September 2018.The GoodReports online tool asks researchers a set of simple questions about their research to "diagnose" which reporting guideline they need. The tool offers an online version of the recommended checklist, with pop-up explanations for each item and examples of good reporting where available. The checklist can also be downloaded to use offline. GoodReports asks authors to go through the checklist, clarify or add more information to their manuscript if required, mark where in the manuscript they have reported the information, and submit the filled checklist with their manuscript. The GoodReports prototype contains 16 reporting guidelines. A pilot in four BMC journals found authors using GoodReports were more likely to use the right reporting guideline. Preliminary data from a study of BMJ Open authors suggested authors using the tool were more likely to submit a reporting checklist with their manuscript than those that don't.We propose to improve GoodReports from prototype to sharable tool. We will add more checklists, make it possible for users to combine elements from more than one reporting guideline, and add at least one example of good reporting for every checklist item. Development will be based on data from our two preliminary studies and user feedback. We will continue to collect GoodReports user feedback online throughout the project to iteratively direct how the tool develops. We will also ask the opinions of early-career researchers in local focus groups and of authors and other potential users (e.g., journal editors and medical writers) in wider online surveys. We will then test whether the GoodReports tool increases the completeness of manuscripts submitted to a medical journal in a randomised trial.