同行评审：如何审稿综述论文？Reviewing review articles, part 2: The how

 by Matt Pavlovich (the editor of Trends in Biotechnology)

Last month, I talked about some of the reasons that you might want to accept that invitation to review a review article. Let's say you found my argument persuasive; now what? The internet (and even this very blog) is full of suggestions of how to peer review scientific research, but advice for how to review reviews is scarce. In fact, the only specific advice for reviewing review articles I could find was this single paragraph from Springer.
How to peer review："A review article is written to summarize the current state of understanding on a topic, and peer reviewing these types of articles requires a slightly different set of criteria compared with empirical articles. Unless it is a systematic review/meta-analysis methods are not important or reported. The quality of a review article can be judged on aspects such as timeliness, the breadth and accuracy of the discussion, and if it indicates the best avenues for future research. The review article should present an unbiased summary of the current understanding of the topic, and therefore the peer reviewer must assess the selection of studies that are cited by the paper. As review article contains a large amount of detailed information, its structure and flow are also important." 

Yes, review articles are peer-reviewed!

综述文章照样需要同行评审！

在发表一篇综述之前，一定会先验证它的科学准确性，并且了解更广泛学界的研究兴趣。
Submitting to a Trends journal requires an affirmative from the editor in some form: either the editor reaches out to a potential author with an invitation to write an article, or the potential author submits a proposal for an article, which the editor chooses to invite for submission. It's a common enough misconception that, because an editor has already given an article (or at least its topic and outline) the green light to submit, reviews don't need to be reviewed and are simply accepted as a matter of course. That's not the case at Trends—we need to verify a manuscript's scientific accuracy and interest to the broader research community, among many other aspects—and it's not the case at many other journals that publish reviews. 

但我认为，让审稿人对评审综述文章一头雾水的主要原因，是不同期刊对综述文章风格和期望的要求“千人千面”，审稿人有时很难把握期刊想要的到底是什么，例如百科全书式详尽的Annual Reviews对综述评审的要求肯定和我们Trends的要求不一样。虽然本文主要讨论Trends的综述评审，但我还是希望其中一些建议对于评审综述文章具有普遍意义。

But I think the vastly different style and expectations of review articles in different journals make it tough to know what any journal in particular is looking for: the encyclopedic and exhaustive Annual Reviews, for instance, surely ask their reviewers to evaluate review articles differently from how we might at Trends. While this post is deliberately Trends-centric, I hope some of the advice is universally applicable to thinking about giving constructive feedback on any sort of review article.

评审意见应该关注什么？What should I comment on?

我们构建了结构化的审稿意见表，引导评审考察稿件的方方面面。我最近对Trends审稿人指南进行了更新，使其更加清晰实用。在本文中，我会我的审稿指导原则作一个比较详细的阐述。（虽然在评审短文章时审稿人不会看到这些，但它们仍然构成了一个可以引导审稿过程的思考框架。）

We structure our reviewer comment form to walk you through the aspects of a potential article that we find most important. I've recently clarified the Trends reviewer guidelines to make them more useful and informative, and I'd like to describe my thought process in a little more detail here. (Even though you won't see these questions if you've been asked to review a short article, they still make up a helpful framework to guide your thoughts.)

中立原则。对于综述文章，审稿人需要明确地回答这篇文章对于其主题是否提供了一个中立地观点。我们试图确保这一主题在当下的所有实际应用、学术观点和理论假设都适当地囊括在内。我们非常不想看到那种作者广泛自我引用，基本重申本人观点的综述。这种文章，名为综述，实为自夸。

但是在评审观点（Opinion）文章时，审稿人不会看到这一原则。这是因为一方面，观点文章理应推动一些议程，并表达个人想法。观点文章并不一定是为了给目前所有观点提供平等展示的平台。恰恰相反，有时它的目的就是推进那些非正统或有争议的想法。但另一方面，观点文章的作者们也不能主观臆测或提出不切实际的范式。

我们对于观点文章的期望，是能够根据现有的文献，判断作者的提议是否合理，并且至少对与之相悖的假设和提议的局限性有初步讨论。有时，审稿人会仅仅因为不同意作者所阐述的观点，而建议拒稿一篇观点文章。我个人并不认为这种意见有很大的建设性。提出观点就是为了激发辩论，如果该领域的每个人都已经达成了共识，那么发表观点文章自然就没有多大价值了。

Balance: For Review articles, you'll explicitly be asked if the article provides a balanced view of the topic. Here, we're trying to ensure that all of the current applications, ideas, and hypotheses are properly accounted for. Something we'd really like to avoid is a review that's glorified self-promotion: one that relies extensively on self-citation and essentially reiterates the authors' own positions from their prior publications.

You won't see this item for Opinion articles. On one hand, Opinion articles should push some agenda and be associated with a personal hypothesis. They're not necessarily designed to give equal platforms to all of the ideas that are out there. Rather, they may be advancing ideas that are unorthodox or controversial. But on the other hand, authors of Opinions don't have carte blanche latitude to speculate wildly or propose unrealistic paradigms.

What we're asking you to do for Opinions is to determine whether the author's proposal is sensible, given the existing literature on the topic, and whether competing hypotheses and the shortcomings of the proposed idea have at least been discussed. Sometimes, we get reviewer comments that recommend rejecting an Opinion article simply because the reviewer disagrees with the opinion that the authors have articulated. I personally don't find this comment especially useful: opinions are meant to stimulate debate, and if everyone in the field already held the same viewpoint, then the article would hardly be worth publishing.

科学准确性原则。顾名思义，我们想知道作者是否正确地解释和展示了相关结果，是否遗漏了任何最新或关键参考文献。Trends期望文章能简明扼要地讨论某一主题过去几年的相关研究。因此，一篇综述通常不可能穷尽地引用关于该主题的每一篇文章。

在这种情况下，审稿人的反馈就显得尤为重要。尽管我们编辑会努力了解期刊所涉及到的所有研究领域，但审稿人才是这个领域的专家。审稿人比我们更清楚，作者是否误读了某个发现，误解了某个术语，或者依赖于某个过时的机制。如果有相关的情况，还请审稿人一定要告诉编辑。需要添加的重要参考文献，或者需要澄清的术语这样的意见也很有帮助。

我们也想听听审稿人对于文章中图表的看法。它们是否容易理解？是否能够补充综述的信息？

Scientific accuracy: This one is more straightforward. We'd like to know if the authors have interpreted and presented the relevant results correctly and if there are any recent and critical references that might be missing. Trends articles are meant to be concise and to focus rather narrowly on research from the past few years, so it is often impossible for an article to cite every single article ever published on the topic.

This is where feedback from our reviewers is especially important! Although we as editors try to stay informed on all of the topics that our journals cover, you're the subject matter expert. If you think an author has misrepresented a particular finding, defined a term incorrectly, or relied on an obsolete proposed mechanism, you know that better than we do, so please let us know. Comments on important references to add or terminology to clarify are helpful too.

We'd also like to hear your thoughts on the manuscript's figures. Are they easy to understand? Do they complement the message of the review?

时效性原则。文章的大部分参考文献应该是过去2~5年的主要研究。当然这不是死规定，不是说2010年的研究或综述就应该从参考文献部分删掉。但是如果一篇综述主要引用的都是10年前的研究，可能意味着这一主题在当前并不是研究兴趣点，本期刊的读者对其关注度可能不高；如果一篇综述只是之前综述文章的简单拼凑，就可能会与主要文献相去甚远，不会有真正独特的观点。

Timeliness: Most of an article's references should be to primary research from the past 2–5 years. Of course, this doesn't mean that any research article from 2010 or introductory review should be ruthlessly stricken from the References section. But a review that mostly refers to research conducted a decade ago might suggest a topic that's not of enough current interest for the journal’s readers to care about. And a review that just pieces together other reviews might be too far removed from the primary literature to provide a truly unique viewpoint.

创新性原则。这可能是Trends综述中最具特色的一个方面。审稿人应该能在评议的稿件中看到新东西，给读者提供在其他地方找不到的信息。一篇综述需要创新到什么程度才可以在Trends上发表呢？其实我们的要求并不高，可以是一个从来没被全面综述过的当今热点话题、也可以是一项技术的新应用、基于新证据的路径更新；提出一套最优实操方法或优值用于新结果的评价，也是很有意思的。

一篇综述如果只列举实验结果，而不给出任何综合性意见、联系或批评，那对读者来说只能算是一个合理的最新参考书目。同样，一篇综述如果只是简单指出 "未来还需更多研究"，那么它就不如提出具体地未来实验展望或商业化发展、临床转化路径建议的综述有趣。

如果最近有与稿件类似的综述发表，也请审稿人告知我们。如果是同一主题的综述，但包含的信息或预期受众不同，也可能还有发表价值。

Novelty: As I discussed last month, this is probably the single most characteristic aspect of Trends reviews. The manuscript you're reviewing should say something different and give a reader information that he couldn't find elsewhere. What makes a manuscript sufficiently novel to publish in a Trends journal? That might be as simple as "This is a topic that people care about now and hasn't been comprehensively reviewed yet." It could also be "Here's a new application for a technology, or an update to a pathway based on new evidence." A proposed set of best practices or unifying figure of merit for how to evaluate new results is certainly interesting as well.

A review that only lists experimental results without providing any synthesis, connection, or critique doesn't give anything useful to its readers beyond a reasonably up-to-date bibliography. Similarly, a review that concludes simply by observing "additional research is needed" is not as interesting as one that posits which experiments should be conducted or suggests a path forward to commercialization or clinical translation. 

Finally, if a similar review has been published lately, let us know. There may still be room for another review on the same topic if it has a different message or intended audience.

权威性原则。多学科特色很强的文章是固然是好的，但有时作者本人其实只专于其中某个特定学科。比如审稿人在审阅一篇关于免疫测定法研究进展的综述时，发现对方法应用的讨论深入且全面，但对测定法所依据的基础免疫学的讨论却很有限，比较混乱。这种情况下，审稿人就可以向编辑建议增加一位与学科背景不同的作者，加强稿件内容。

Authoritativeness: Strongly multidisciplinary articles are wonderful, but sometimes the authors' primary expertise is in just one discipline. Let's say you're reviewing a review on an immunoassay that has emerged in the past couple of years, and you notice that the discussion of the applications is critical and comprehensive, but the basic immunology underlying the assay is limited and confused. This situation would be a great time to suggest to the editor that an additional author, with a different background from that of the original authors, could strengthen the manuscript.

易读性原则。我们明白作为专家，你在这个领域浸淫已久。但在审稿时，我们希望审稿人能退一步去看，想象一下如果你没有发表过几十篇关于这个主题的论文，还能看懂这篇稿子吗？我们希望审稿人能对文章结构是否合适、主要信息（以及作者撰写综述的初衷）是否清晰，缩写和缩略语是否有用和标准等方面，发表您的意见。

Accessibility: We understand that, as an expert, you're immersed in the field, but here, we'd like you to take a step back and imagine that you haven't published dozens of papers on the topic. Would you still be able to understand the manuscript? We want you to comment on the suitability of the structure, the clarity of the take-home message (and the justification for writing the review in the first place), and whether abbreviations and acronyms are useful and standard or just confusing.

审稿意见不需要关注什么？ Is there anything I don't need to comment on?

每隔一段时间，我就会看到有些审稿意见，对稿件内容的批评只占全文的三分之一，剩下三分之二全是对措辞、语法和标点符号的建议。如果作为审稿人的你，其实有一颗想当编辑的心，并且很想在审稿时实现自己的这一“愿望”，那我们编辑举双手赞成。

但我认为审稿人不应该觉得自己有义务在文字上着力。每篇Trends文章都会在预审过程中得到个性化编辑意见。我们会特别关注稿件语言是否清晰明了，是否适合本刊读者，之后建议作者重写某段或对某些章节重新编排。一旦稿件被确认接收，就会进行专业的文字编辑和排版，保证每一个标点符号都会在发表时出现在正确位置。

同样，也请审稿人不要担心参考文献的格式、数量、字数。但如果你认为某一节（或整篇稿件）太长或太短，这对我们来说是很有用的。

最后我想说，有时需要评议的文章可能会和审稿人本人的想法大相径庭。我们需要审稿人根据一篇文章的独立价值来评估它，而不是将文章与您本人会提出的论点或您会选择纳入综述的文献进行比较（如果您持有不同意见，Trends编辑很乐意让您以来信（Letter）的形式对文章作出回应）。

Every once in a while, I get comments from a reviewer that are one-third critique on the content and two-thirds suggestions for usage, grammar, and punctuation. If you missed your calling as a copyeditor and love commenting on these elements, great! It makes our jobs easier. 

But I want to assure you that you shouldn't feel obligated to comment on these aspects. Every Trends article gets independent editorial comments in a process that we call pre-revision. We focus in particular on the manuscript's clarity and suitability for the journal's audience, and we suggest passages for our authors to re-write or sections to re-structure. Then, once a manuscript is accepted, it's professionally copyedited and typeset, which means that every comma will be in its right place upon publication.

Similarly, feel free not to worry about formatting, number of references, or word count—though if you think a section (or entire manuscript) is too long or short, that's useful for us to know.

Finally, sometimes the paper that the authors wrote is not the one you would have written. We ask that you evaluate an article on its independent merit rather than comparing it against the arguments that you would have made or the canon of literature you would have chosen to review. (If your perspective diverges from that of the authors, I and many of the other Trends editors would be delighted to have you respond to the article in the form of a Letter.)

综上所述，审稿人应该给出怎样的评议结论？Putting it all together: What should I recommend?

审稿人会被要求对稿件给出一个评议结论，是接受、小修、大修（不需要额外审阅）、大修后重审、还是拒稿。在考虑了我上面所说的所有关键方面之后，怎么才能知道该给出哪个结论呢？

这里我想介绍一个“三点评估法”。想象一下，一个对本期刊领域感兴趣，但对本文章所讲内容了解不深的普通研究人员，

1. 能够理解这篇综述吗？

2. 能够从中了解准确的、最新的研究概况吗？

3. 能够从中知道一些在其他文章中读不到的东西吗？ 

尽管有些综述可能需要修改，需要打磨语言、理清结构、增加额外的参考文献，但如果文章能经得起这种评估，那就是值得发表的。

You'll be asked to provide a recommendation: accept, minor revision, major revision (without additional review), major revision and re-review, or reject. But after considering all of the key aspects I've described above, how do you know which one to pick? 

I propose a three-pronged test. Imagine an average researcher who is interested in the areas covered by the journal that you are reviewing for but is not an expert in the specific topic of the review paper.

First, can she understand the review?

Second, does the review give her an accurate and current representation of the scientific topic?

And third, will she learn something unique from this review that she wouldn't learn by reading another article?

A review manuscript that survives this scrutiny is worth publishing, albeit possibly with some revisions to clarify the language, tidy up the structure, or incorporate additional references.

为了更加清晰，审稿人也可以这样想：

“接受”，意味着稿件完全符合这些标准，只有改进语言或格式就好。

“小修”，意味着稿件基本符合这些标准，但仍可通过更好的数据、更多的讨论、传递更明确地信息或技术澄清来改进。

“大修”，是指稿件目前不符合标准，但一旦作者有机会修改，就有很大机会符合标准。

“大修加重审”，是指稿件目前不符合标准，但主题足够有趣，你想再给作者一次机会，让他们提交符合标准的东西。

“拒稿”，则是该稿件有严重的缺陷，即使经过大量的修改也不可能符合标准。

For even more clarity, you might think of it this way:

• Accept means that a manuscript meets those criteria as-is, and the only improvements that should be made are to the language or formatting;

• Minor revision means that a manuscript meets those criteria but could nevertheless be improved with better figures, additional discussion, a stronger message, or technical clarifications;

• Major revision means that a manuscript currently doesn't meet the criteria but has a good chance to once the authors have had the chance to revise it;

• Major revision with re-review means that a manuscript doesn't currently meet the criteria, but the topic is interesting enough that you'd like to give the authors the chance to come up with something that does;

• And reject means that the manuscript is so deeply flawed that it is unlikely to meet the criteria even after extensive revision.

I hope that these comments are useful—and that they've inspired you to consider reviewing manuscripts for the Trends journals. And if there are any other suggestions for reviewing review articles out there, we would love to hear your tips too.

Information for Trends reviewers

What should I comment on?
The first section of the reviewer report asks a series of yes/no questions. These questions ask about some of the most important aspects of Trends articles and are designed to guide your thoughts on the manuscript. Your answers are not made available to the author. Where applicable, we encourage you to further explain your answers to these questions in your comments.

• Balance: Are readers likely to come away from the article with an accurate idea of the current state of the field, or would they benefit from inclusion of additional work, alternative theories, etc.? Do the authors dwell excessively on their own work? (You will only see this item for Review articles, but please consider these points for articles in any format.)

• Scientific accuracy: Have all the facts been presented clearly and precisely? Are all of the most crucial recent references included and cited? Are the figures accurate and easy to follow?

• Timeliness: Is there a compelling reason to publish on the topic at this time, and have the authors made that reason clear? Do most of the references date from the past 2-5 years, or is there too much reliance on older references? For references, is there too much reliance on reviews instead of original research?

• Novelty: Does the manuscript offer readers something new? Is there some unique insight or interpretation, or is the manuscript simply a list of results?

• Authoritativeness: Are the authors qualified to comment on this topic? Would the manuscript benefit from another author with greater expertise or a background in a different area?

• Accessibility: Will non-expert readers and students understand the main message? Is the manuscript structured in a way that makes sense?

The editor will independently comment on other important aspects of the manuscript, such as language, clarity, and adherence to the journal's formatting requirements. Furthermore, all Trends articles undergo professional typesetting and copyediting prior to publication, so your review can focus on the concepts listed above rather than these details.

参考资料：https://www.cell.com/trends/reviewers