|
How to Review a Clinical Research Paper 如何评阅临床研究论文
Michael D. Hill 迈克尔·D·希尔
Peer review is an essential component ofthe scientific process. It is imperfect, to be sure, but there is widespreadagreement that it is the best way to ensure that reliable scientificinformation is published. Being a reviewer is only 1 component of the processof publication. If you are an author or want to be an author, you have a dutyto take part in reviewing your colleagues’ papers, just as your colleagues havereviewed your papers. Reviewing papers is a helpful part of learning thetechnical art of medical writing because you see and learn by example, bothgood and bad. Although it is a volunteer duty, there is a skill in providing auseful review and mentorship and experience matter in how you provide yourreview. Herein, I provide some steps on how to review papers for Stroke,specifically focussing on clinical papers. 同行评阅是科学过程的重要组成部分。虽然它并不完美，但人们普遍认为，这是确保发表可靠的科学信息的最佳途径。审稿只是发表过程的一个组成部分。如果您是作者或希望成为作者，那您有义务参与评阅您同行的论文，就像您的同行评阅您的论文一样。评阅论文是学习医学写作的技艺的有用部分，因为您通过实例见识和学习，高下立见。尽管这是一项志愿者义务，仍有一些技巧、指导和经验有助于提供有用的评阅意见。在此，我提供了一些关于如何评阅《Stroke》的论文、特别是临床研究论文的步骤。
Editorial Structure at a Journal 期刊编辑结构
At many journals, including Stroke, thereare a group of associate editors who handle papers through the review andpublication process. These people are typically your senior peers and are alsovolunteers. Your job and duty as a reviewer is that you are advisory to them.The associate editors look at each paper as it comes in, and they may reject apaper outright or provide comments back to the author group for revision evenbefore a paper is sent for peer review. In sending a paper to you as areviewer, the associate editors seek your advice on quality, content, andcontext. Comments that you make to the editors directly are confidential andoften helpful to them to interpret the context of a given paper. 在包括《Stroke》在内的许多期刊中，有一批副编辑在评阅和发表过程中处理论文。这些人通常是您的高年资同行，也是志愿者。您作为审稿人的工作和职责是您给他们提供咨询。副编辑审阅每篇文章的内容，他们可能会直接拒绝一篇论文，或者甚至在论文发送给同行评阅之前向作者提交意见。在向您发送论文、请您担任审稿人时，副编辑会在质量、内容和语境方面寻求您的建议。您直接向编辑提出的评阅意见是保密的，并且通常有助于他们解读这篇论文的语境。
Decide Whether You Are Qualified,Interested, and Have the Relevant Expertise. Recuse Yourself if You Have aConflict of Interest 判断您是否合格、有兴趣、并具有相关的专业知识。如果您有利益冲突，请回避。
Read the title and abstract of the paperand decide whether you have the relevant expertise and interest to provide areview. If the topic is not in your wheelhouse, decline the review. Similarly,be sure you can dedicate the time to do this well. It may take 30 minutes or itmay take 3 hours or longer to review a paper well. If you cannot dedicate thetime, decline the review. Review the list of authors and decide whether youhave a conflict of interest. If you are working closely with ≥1 of the authorsor have recently published with them, you may have a conflict of interest, andyou should decline the review. If you have a financial interest in the topic athand, the sponsor of the study, competing intellectual property, then you mayhave a conflict of interest. If you have doubt about possible conflicts,contact the handling editor and discuss it with them. Try to complete thisinitial task within 48 hours of receipt of the request to review so that theassociate editor can search for other reviewers in the event that you declinethe task. 阅读论文的标题和摘要，并确定您是否有相关的专业知识和兴趣来提供评阅。如果话题不在您的掌控之中，请拒绝评阅。同样，确保您可以花时间做好这件事。可能需要30分钟或者3个小时或更长时间才能完成评阅。如果您不能投入时间，就拒绝评阅。查看作者列表并确定您是否存在利益冲突。如果您与其中一名或更多作者密切合作或最近与他们合作发表过，则可能存在利益冲突，您应拒绝评阅。如果您与手头的主题有经济利益关系，即研究的赞助商、知识产权竞争者，那么您可能存在利益冲突。如果您对可能的冲突有疑问，请联系责任编辑并与他们讨论。尽量在收到评阅请求后的48小时内完成此初始任务，以便副编辑可以在您拒绝任务时寻找其他审稿人。
Review With a Mentor or Content Expert. LearningFrom Other Reviewers 与导师或专家一起评阅。向其他审稿人学习。
If you have limited experience withreviewing papers, review with a mentor. Often your mentor may ask you orintroduce you to the review process by asking you to do a review. You can learnfrom the final review submitted. In addition, Stroke will share with you thereview comments from other reviewers. Read them so that you can learn fromthem. 如果您评阅论文的经验有限，请与导师一起评阅。通常，您的导师可能会通过要求您审稿来使您参与评阅过程。您可以从提交的最终评阅意见中学习。另外，《Stroke》将与您分享其他审稿人的评阅意见。阅读它们，您可以从中获得借鉴。
Review the Paper 评阅论文
Do your utmost to stick to the timelinesrequested by the journal. As an author, you know how frustrating it is to bewaiting for reviews. Prompt reviews help the overall process of science. Reador skim the paper through completely once so that you understand the overallpaper and results. Then go back and read in detail. Look for the following characteristics: 尽最大努力在期刊要求的时间内完成。作为一名作者，您知道等待审稿有多令人沮丧。及时的评阅有助于推动整个科学过程。概览论文，以便了解整篇论文和结果。然后再仔细阅读。查看以下方面：
Content 内容
1. Look for a concise review of thebackground around the topic. Do you agree with it? Has anything been missedthat you know of? Does the introduction set the context for the work? Does theintroduction end with a statement of purpose or aim for the study that is beingreported? Are the key papers in the field noted and referenced? 1.查看围绕主题背景的简要综述。您同意吗？您知道有什么东西被遗漏吗？引言是否为研究设定了语境？引言是否以研究目的结束，或提出进行研究的目的？该领域的关键论文是否被提及和引用？
2. Read the Methods and focus on them themost. The principle is the same for all types (clinical or foundationalscience) of scientific reports: the Methods should have enough detail that ifthe study was repeated using these methods, the same essential result(s) couldbe obtained. Read any supplementary methods that may be included as ane-appendix. Is the population studied clearly identified and characterized? Arethere clearly stated primary outcome, secondary outcome, and exploratoryoutcome statements? Are the statistics clear and well described? Are theinclusion and exclusion criteria clearly stated? What are the details of thedata collection? How were the data managed? What was the process of statisticalanalysis, and who performed the analysis? If you do not have statisticaltraining and cannot interpret complex statistics, it may be relevant torecommend to the handling editor that a formal statistical review be conducted.For specific study types, follow accepted checklists. For example, randomizedcontrolled trials, meta-analyses of trials, meta-analyses of cohort studies,and others have standardized guidance on what should be included in a paper.These checklists, such as the CONSORT statement (http://www./)and PRISMA guidelines (http://www./), are helpful inassessing the completeness and appropriateness of reporting. For these specificstudy types, each guideline requirement is required; it is not optional forauthors not to include certain components at their discretion. 2.阅读“方法“并给予最高重视。对于科学报告的所有类型（临床或基础科学）来说，原则是相同的：方法应该有足够的细节；如果使用这些方法重复研究，则可以获得相同的基本结果。阅读可能包含在电子附录中的任何补充方法。研究的人群是否明确确定和表征？是否有明确的主要结果、次要结果和探索性结果陈述？统计数据是否清晰并有详细描述？纳入和排除标准是否明确说明？数据收集的细节是什么？数据如何管理？统计分析的过程是什么，谁完成了分析？如果您没有经过统计培训并且无法解释复杂的统计数据，则可能需要向责任编辑建议进行正式的统计评阅。对于特定的研究类型，请遵循公认的清单。例如，随机对照试验、临床试验的荟萃分析、队列研究的荟萃分析以及其他已经对应该包含在论文中的内容有标准化规定的研究。这些清单[如CONSORT声明（http://www./）和PRISMA准则（http://www./）]有助于评估研究报告的完整性和适宜性。对于这些特定的研究类型，每个准则要求是必需达到的;作者不可自行选择包含某些部分。
3. Clinical ethics is an importantconsideration. Human research ethics deals with interventions on humans, theirdata, and their tissues or cells. Look for a statement that indicates that thestudy was reviewed and approved by the relevant ethics board. 3.临床伦理是一个重要的考虑因素。人类研究伦理涉及对人类、人类数据、人类组织或细胞的干预。查看表明该研究已经由相关伦理委员会审查并批准的声明。
4. The results should be concise and aresupplemented with key figures and tables. The results should reflect themethods in organization and structure. The text should not repeat what is inthe tables and figures. This latter phenomenon is common and can be rectifiedwith simple editing. 4.结果应简明扼要，并附上关键的图和表格。结果应该按方法的组织和结构展示。文字不应重复表格和图中的内容。后一种现象很普遍，可以通过简单的编辑来纠正。
5. The Discussion and Conclusions shouldstick to the data being presented. Some authors will overreach, stray away fromtheir data specifically, and make broader generalized conclusions; these shouldbe saved for the editorialist and not included in the main paper. Theconclusions should directly reflect the data and findings. The Discussionsection should also include a paragraph describing the known limitations of thedata, the analysis, and the conclusions. 5.讨论和结论应该围绕所提供的数据。一些作者会超越，特别是偏离他们的数据，并作出更广泛的普遍结论。这些应该保留给编辑，并且不应包含在主要论文中。结论应直接反映数据和结果。讨论部分还应描述数据的已知局限性、分析和结论。
6. Review the references. Are keyreferences missing? 6.查看参考文献。是否缺少关键参考文献？
7. Look for evidence of plagiarism. You maynotice that content is identical to another paper based on your knowledge ofthe field and other recent papers. You may notice a sudden change in style orquality of writing from one paragraph to another. You can copy and paste a sectionof the paper into an internet search engine to see whether it can be recognizedelsewhere. Some journals will have electronic search capability to detectcopied sections, and the journal editorial staff will look for this. You canaid this process with your expert knowledge of the field and its literature. Ifyou detect plagiarism or think you do, mention this observation to the editor inthe confidential comments to the editor section. 7.寻找剽窃的证据。您可能会发现，根据您对该领域和其他近期论文的了解，某篇论文内容与另一篇论文完全相同。您可能会发现，从一个段落到另一个段落的写作风格或写作质量的突然变化。您可以将一部分论文复制并粘贴到互联网搜索引擎中，查看它是否可以在其他地方搜索到。一些期刊将用电子检索功能来检测复制的部分，期刊编辑人员会做这些事。您可以借助您对该领域及其文献的专业知识来帮助完成此过程。如果您发现剽窃或者认为可能剽窃，请在给编辑部门的保密评阅意见中提及。
Common Stylistic Considerations to Commenton 常见写作风格问题
1. Look for cliché language and reject it.For example, “the only treatments for acute ischemicstroke are intravenous tissue-type plasminogen activator and endovascularthrombectomy”…. You can assume that the audience that reads Stroke knows this,and the sentence can be deleted. Not only was Shakespeare correct in notingthat “brevity is the soul of wit,” copying someone else’s turn of phrase can beconsidered plagiarism and is, simply put, unoriginal. 1.寻找陈词滥调并拒绝它。例如，“急性缺血性卒中的唯一治疗方法是静脉内组织型纤溶酶原激活剂和血管内血栓切除术”......。您可以假设《stroke》读者都知道这一点，并且删除这一句。不仅莎士比亚指出“言以简为贵”，而且抄袭别人的短语可以被认为是剽窃。简而言之，它是非原创的。
2. Grammar and style matter: if English isa second, third, or more distant language for the authors, grammar and stylediffer. The manner of expressing ideas varies by language, and you can see bydirect translation of certain phrases that the same concept is expressed in astylistically different way in various languages. Journals have greater orlesser degrees of copyediting resources, and so some simple comments from you thereviewer can be helpful. Encourage authors to use available electronicresources, such as a spell checker and grammar checker; these are available inMicrosoft Word and other software packages. But beware, spell checkers andgrammar checkers are also not perfect and will not recognize a correctlyspelled word that has the wrong meaning. 2.语法和写作风格问题：如果作者的英语是第二、第三或更远的语言，则语法和写作风格不同。表达观点的方式因语言而异。您可以通过直接翻译某些短语来看到相同的概念以各种语言、各种写作风格表达。期刊拥有或多或少的润色资源，因此审稿人的一些简单评阅意见可能会有所帮助。鼓励作者使用可用的电子资源，如拼写检查器和语法检查器，这些都可以在Microsoft Word和其他软件包中找到。但要小心，拼写检查器和语法检查器也不完美，不会识别含义错误而拼写正确的单词。
3. Stylistically, there are many phrasesthat are useless—“statistically significant,” “respectively,” “this is thefirst study to show”…. If you are commenting on a finding that is statisticallysupported by a hypothesis test, then it will be clear from the tabulatedresults with supporting P values or confidence intervals. The word respectivelycan lead to confusion; state clearly what you are referring to. Finally,claiming primacy of the observations is immodest, unhelpful, and the authorscan never truly know whether they were first. These can all be removed and willsave printed space. Do not by hypercritical when a paper is obviously writtenby a non-English speaker; just advise the authors to have a native English speakerhelp them with the copy editing and phraseology. 3.在写作风格上，有许多短语是无用的- “具有统计意义”，“分别”，“这是第一个研究显示”......。如果您对一项由假设检验统计支持的结果发表评阅意见，那么从支持P值或置信区间的表格结果中可以清楚地看出。“分别”这个词会导致混乱。要明确说明您所指的是什么。最后，声明观察的首次地位不够谦虚，无益。作者永远不会真正知道他们是否是第一个。这些都可以删除，节省印刷空间。当一篇论文明显是由非英语作者撰写的时候，不要过分苛求；只是建议作者让一位英语母语人士帮助他们进行润色和措辞。
Comments to the Authors 给作者的评阅意见
1. I recommend numbering your comments andproviding 1 major comment per number. The authors will be asked to respondpoint by point to your reviews if the editor recommends revisions. Numberingthe results helps the authors. 1.我建议将您的意见逐条列出，一个重要评阅意见为一条。如果编辑建议修改，作者将被要求逐条回复您的评阅意见。将评阅意见编号有助于作者。
2. Try to be both critical andconstructive. In part, your job is to help the authors make the paper better.Constructive comments telling the authors how to improve their work will onlyhelp. 2.尽量做到既有批评性又有建设性。部分原因是，您的工作是帮助作者更好地完成论文。建设性的评阅意见告诉作者如何改进他们的工作。
3. If the paper is reporting a morestraightforward observation, encourage the short report format; in Stroke, thismeans a 1800-word paper. This takes nothing away from the paper or observation,and such a publication is listed in PubMed just as any other paper. Most authorsare verbose in their writing. Tables tell a lot of a story and much materialcan be reported in a table or put into an e-appendix. 3.如果论文报告的是更直接的观察，请鼓励使用短篇报告格式。在《Stroke》中，这意味着一篇1800字的论文。这不会对论文或观察产生任何影响，这样的出版物就像其他任何论文一样在PubMed中列出。大多数作者在写作时都很冗长。表格能讲述很多故事，许多材料可以列在表格中报告或放入电子附录中。
4. Finally, make a call and be definitive.Your judgment matters, and you should decide whether a paper is or is notuseful to the field. There is a tendency to choose a middle of the roadresponse—major revisions—even if the paper really is not that good or thatuseful. A paper may be extremely well written and presented but still be lowimpact. Maintain a high standard, and if the paper does not meet that standard,reject it. In contrast, if the paper is good, then say so to the editor. Majorrevisions can still apply to a good paper. 4. 最后，决断公正，态度明确。您的判断很重要，您应该决定一篇论文是否对该领域有用。有一种趋势是回复时采取中间道路– 重大修改 - 即使这篇论文确实不太好或有用。一篇论文有可能写得非常好，但仍然价值不高。保持高标准！如果论文不符合该标准，请拒收。相反，如果论文不错，那么就向编辑说。重大修改更适用于一篇好论文。
5. Remember you are giving advice to theeditor. Provide confidential comments to the editor in the space provided togive your critique some context. Tell the editor (but not the authors) whetheryou think the work deserves publication or not. It is the editor’s job to readthe paper and make a decision given your advice. Do not be upset if the editormakes a different decision than you. It is their prerogative as the editor. 5.记得您正在给编辑提供建议。在指定空间中向编辑提供保密评阅意见，以便为您的意见提供一些语境。告诉编辑（但不是作者），您认为该论文是否值得发表。阅读论文并根据您的建议做出决定是编辑的工作。如果编辑做出的决定与您的建议不同，别不高兴。这是他们作为编辑的特权。
After Your Review 在您评阅之后
Consider registering your review with aservice, such as Publons at www.publons.com. This service keeps track of yourreviews in an online database. These data can be useful for you if you are anacademic and can be reported at your University for academic credit. When thejournal editor has completed the overall paper review, they will typically sendyou a summary of their decision and the other reviewers’ comments. Read themand consider them in context because there may be learning points from yourcolleagues’ reviews. Finally, maintain confidentiality. Your knowledge of thepaper and its review is privileged information that should not be shared. 考虑使用在线服务注册您的评阅，例如Publons，网址www.publons.com。此服务会在网络数据库中跟踪您的评阅。如果您是学生，这些数据可能对您有用，并且可以向您的大学申请学分。当期刊编辑完成整篇论文评阅时，他们通常会向您发送他们的决定摘要和其他审稿人的评阅意见。阅读它们并在语境中琢磨，可能会从同行的评阅意见中学到一些。最后，保持机密。论文及其评阅意见是不应分享的私密信息。
Conclusions 结论
Reviewing clinical papers is an importantand useful part of your early medical career. You will gain insight into the currentliterature, learn about medical writing, and by exposure, gain a betterunderstanding of the limitations of medical evidence and clinical research.When done well and with care, peer review can make the reporting of clinicalscience better and more transparent. 评阅临床论文是您早期医学生涯的一个重要且有用的部分。您将深入了解当前的文献，了解医学写作，并通过接触评阅意见，更好地了解医学证据和临床研究的局限性。如果做得好且认真，同行评阅可以使临床科学的报告更好、更透明。 原文： Hill MD. How to Review a Clinical Research Paper. Stroke, 2018,49(5):e204-e206. |
|
|
