Guidance from our Editors

Here is some practical advice on how to conduct a good review from the point of view of our experienced Editors.

Emad El-Omar, Editor-in-Chief of Gut


  1. Start by providing a succinct summary about the work and the researchers. Editors like to know the standing of the group in question and the quality of their work.
  2. Novelty: this is very crucial. The Editor wants to know how novel the work is and what is novel about it. Has the work answered a new question or did it merely corroborate published work? Does the reviewer know of similar work that has just been published. Has the reviewer reviewed the work for another journal and have the authors made any changes to the previous version of the paper?
  3. Be as critical as you can in relation to the design of the study. Provide the killer blows that would sink the work but only if you can back them with evidence. If the work is flawed but salvageable, tell us what needs to be done to rescue the paper. This is helpful to the Editor, the authors and to science in general. Avoid vague criticisms that are difficult to address or respond to.
  4. List your comments in two groups: major and minor. Major points are ones that require additional experiments, analyses, rewriting etc. Those are ones that would enhance the scientific quality of the work. Minor points are linguistic, grammar, typographical, illustrations etc and would enhance the presentation of the work. It is always helpful to number your comments to allow authors to respond in an orderly manner and editors to track this easily. Generally speaking, it is not necessary to list all typographical errors etc. Simply mention that the manuscript requires careful editing and that errors must be corrected.
  5. Use diplomatic language throughout. You can be very critical but avoid insulting the authors. The aim is to be helpful and to enhance the work overall.
  6. In your confidential comments to the Editor, be as candid as you can regarding the work, highlighting any irregularities, concerns etc. It is also helpful to provide what you think is the minimum requirement for accepting a revised manuscript, even if you have suggested the authors do much more.

Deborah Bowman, former Editor-in-Chief of Medical Humanities


Without reviews and reviewers, the journal wouldn’t exist. Everyone – authors, editors, researchers, publishers and readers – benefits from the contribution of reviewers. We are grateful to you all.
Whether you’re a longstanding reviewer or new to the role, we hope these guidelines might be helpful when you next receive that review request via ScholarOne.

Before the Review
  1. The best way to ensure that we only bother you with requests to review papers that are of interest and within your area of expertise is to update your entry on our reviewer database. If the keywords don’t reflect your specialism, please contact editorial office so we can include your preferred terms.
  2. We know that potential reviewers have many calls on their time and we understand that sometimes you won’t have the time to act as a reviewer. If you’re busy or the review is for a paper beyond your expertise or area of interest, a prompt response declining the review invitation is immensely helpful. It is a bonus if you can recommend a name or two to act in your stead.
  3. If you do accept the invitation to review, please make a note of the deadline when your review is due. If you’re unable to meet the deadline, do get in touch and let us know so we don’t irritate you with reminders.
The Review

A helpful review is:

  1. Constructive – a review that follows the principles of effective feedback and is respectful in tone, but honest about content, is invaluable.
  2. Specific – feedback that is vague or unsubstantiated is of limited use. Ideally, a reviewer will engage directly with the paper, citing specific aspects of the submission or examples from/references to the literature, to back up opinion or commentary.
  3. Aware of the aims and readership of the journal – being alert to the remit and audience of the title and providing comment on the interest and relevance of a submission to the journal’s aims and audience.
  4. Developmental – as well as being specific, comments that articulate what could be improved and how (in broad terms) are much-appreciated by authors and contribute to the overall quality of scholarship and research.
  5. Structured – for clarity, reviews should be organised for ease of reference e.g. by a system of numbering for separate comments. Some reviewers also theme their feedback e.g. general observations, specific comments, opinion as to the originality or importance of a submission etc.
  6. Comprehensive – it attends to every aspect of the submission from the abstract to the conclusion and considers the coherence and fluency of the whole. E.g. are the methods appropriate? How does the theoretical position inform the discussion? Are the questions posed at the outset addressed? Does the conclusion reflect the findings and/or arguments made? What are the limitations of this submission and does the author acknowledge those limitations?
  7. Attentive – is the paper well-written? You are not expected to edit the submission, but if it requires work or attention to spelling, syntax and grammar, a note in the review to this effect is helpful.
  8. Logical – is the reason for your recommendation regarding the submission reasoned and rational based on your comments? It is rare that a paper cannot be improved at all, but if you do feel that the submission should be accepted for publication unchanged, please explain your recommendation. Likewise, if you believe a paper is not likely to reach publishable standard, even with revision, please explain why.
After the Review
  1. When you’ve submitted your review, the Editorial team considers it in conjunction with the other reviews we’ve commissioned to make a decision on the manuscript. If we decide that the authors will be invited to amend their paper and resubmit it to the journal, we will usually share your feedback – apart from any confidential comments to the Editor – with the authors. We always ask authors to demonstrate explicitly how they have engaged with and/or addressed each of the points made by the review.
  2. Sometimes we will invite you to comment on the revised submission. Although the principles of good reviewing remain the same with revised submissions, it is particularly useful at this stage to have your comments on the effectiveness with which the authors have addressed the feedback and comments made in the original peer review.
General Comments on Reviewing
  1. All submissions seen by reviewers should be considered and treated as confidential.
  2. Conflicts of interest may exist, particularly in specialised or relatively small academic fields and reviewers should always consider whether they are able to provide a fair review. For excellent guidance of conflicts of interest, see the guidelines published on
  3. Reviewing should never be delegated or submitted on behalf of another named individual. Reviewers are responsible and accountable for the reviews submitted.
  4. We are always pleased to support reviewers. Please don’t hesitate to get in touch with any concerns, questions or feedback on the experience of being a reviewer. You can contact the Editorial team at the journal directly.

Ian Wacogne, Editor-in-Chief of ADC Education & Practice


Thank you for thinking about reviewing, or agreeing to review a paper for us.

There are very many guides and structures available online for reviewing papers. We’d like to offer you a few tips as part of this.

First and foremost, we’d like to emphasise that your role as a reviewer is to help make things better for patients. This might seem slightly distant from reviewing a journal paper, but it is one of the core values of BMJ Publishing: Patients come first. Your duty, as a reviewer, is ultimately to patients. Think: If this paper is published, and readers act on it, will it improve patient care? Of course, if the paper is not “True” – for example, the science is poor, or the conclusions are false, then it is reasonable to assume that patients would be harmed as a consequence of publication.

This leads quite easily to the second duty, which is to the readership. The people who read the article after publication will trust that it has been carefully scrutinised, and may make changes to their practice based on what they read. Your task is to perform that scrutiny.

There are some other tasks which fall out of the duty to the readership. For example, it is important that a paper has a clear message, is understandable and can be followed by the average reader. It is also important that the message is true – ie not exaggerated, but also not understated.

The third duty is to the authors. They are likely to be similar people to you, who have worked hard to produce their manuscript. As with the duties above, a good review should improve the work, but while it is important to be robust, it is also important to be kind and polite, to explain what parts you disliked about their paper and to suggest improvements. Note that some journals will automatically reveal your identity to the authors; others will give you the option. We’d like to suggest that even if your identity remains concealed, behave – and write – as if it might not. That’s not intended as a threat, and if the journal has agreed to not reveal your identity, you should take their word for it, but you shouldn’t write things that you would not be prepared to say to somebody in person.

The last duty is to the editor and the journal. If you’ve fulfilled the duties above, then you’ll have usually done what the editor needed. However, sometimes you might note that you’ve not been entirely definitive in your opinion. Can we, therefore, politely, ask you to give a clear view of whether you think that the paper should be published, and, if you feel it should, under what circumstances.

Some things you should not do:
  1. Some reviewers write very positive comments to the author, and then put comments for the editor only which are very negative. This is hard to manage, and takes a lot of editorial time in handling the author.
  2. You shouldn’t share the manuscript. If you wish to share it with, for example, a doctor in training as part of an exercise for them, under your supervision, to develop reviewing skills, then that’s acceptable, but you should keep control of the manuscript, and make it clear to the editor that this is what has happened.
  3. Please don’t agree to review if you’re not going to achieve it in the timescale. Authors rely on us to make rapid decisions on their manuscripts, on which may depend their next grant application or their assessment by their governing body.
  4. Please don’t exceed your skills or ability. If you’re worried about the statistics, and are not a statistical expert, then recommend that the editor gets a statistical review. Similarly, if there are aspects of the paper which are outside your skills or experience then please let the editor know.
  5. You don’t need to undertake an extensive copy edit. We will try not to send you badly spelt, punctuated and laid out manuscripts, but if we do and it is unreadable, please let us know. We are asking you to give us an opinion about the “truth” of the manuscript and its conclusions; we recognise that you are busy and therefore we have other processes which will pick up the spelling mistakes. You should only comment in your review on syntax, spelling or punctuation where it distorts the message – either obscuring it or, worse, misrepresenting it. Feel free to mention to the editor if you think the manuscript might need a careful copy edit.