Stetleaf

How to write a constructive peer review

Stetleaf Academy

Reviewed by Alex Juusela, MD, MPH, FACOG, Maternal-Fetal Medicine Physician · Last reviewed 2026-11-05

Peer review works best when the report helps two people at once: the editor, who has to decide what happens next, and the authors, who have to improve the manuscript. A constructive review is specific, fair and organized, so that every point can be acted on. This article gives you a structure you can use for your first review and keep using after that.

Why it matters

Peer review asks experts in the same field to scrutinize a piece of work before it is published. It is a central part of scientific publishing, and it is also criticized for being slow and open to bias [1]. How reviewers write matters to both sides of that picture.

Research on peer review gives two practical lessons. In a study of 93 reports of randomized trials published in open peer review journals, many reports were still missing important methods details after review, so reviewers had not picked them up. Most changes the reviewers did request improved the reporting, but some, such as asking for extra analyses that had not been planned, could make the final report worse [2]. In a randomized trial at one journal, adding a review that looked specifically for missing reporting-guideline items gave a modest improvement in manuscript quality, smaller than the investigators had expected [3]. The lesson for reviewers: check the reporting carefully, and think twice before asking for new analyses.

Key ideas

Before you start

Agree to review only if the topic is within your expertise and you can meet the deadline. Declare any competing interests to the editor, and decline if they are significant. The manuscript is not yours: keep it private, don't share it or discuss it with others without the editor's permission, and don't use its content for your own work [4,5]. Check the journal's policy before using any software, including AI tools, with a manuscript, because many journals restrict this.

Read twice

Read the whole manuscript once for the big picture: what question it asks, how it answers it, and what it claims. Then read it again section by section, making notes as you go. Provenzale and Stanley's step-by-step guide includes a template that works as a checklist of questions to ask [6].

Structure the report

A clear report usually has four parts.

  1. Summary. Two or three sentences, in your own words, on the question, the design and the main claim. This shows the authors and editor that you have understood the paper, and it lets them spot a misunderstanding early.
  2. Major points. Issues that affect whether the conclusions follow from what is reported: a missing description of how participants were selected, an outcome that changed from the protocol, conclusions that go beyond the design.
  3. Minor points. Clarity, missing details, inconsistencies between numbers, and figures or tables that could be clearer. Group small language points together rather than listing every one.
  4. Comments for the editor only. Most journals give you a separate box for these. Use it for concerns you want the editor to weigh, and for your recommendation if the journal asks for one. The decision belongs to the editor.

Make each point actionable

A useful comment says where the problem is, what it is, why it matters, and what would help. Number your points so the authors can answer each one.

Use the reporting guideline

Check the manuscript against the reporting guideline for its design. If an item isn't reported, say which item and what is missing. This keeps your review focused on what readers need, rather than on how you would have written the paper.

Mind the tone

Write about the manuscript, not the authors. Acknowledge what is done well. Avoid sarcasm and sweeping judgments. Write the review you would want to receive.

Worked example

Illustrative example (invented). Not a real manuscript.

You are reviewing an invented cohort study on a postpartum text-message reminder program. Here are three first-draft comments and constructive rewrites.

Comment 1. First draft: "The methods are poorly written." Rewrite: "Methods, participants paragraph: please state how patients were identified in the records (for example, which diagnosis codes were used) and the dates of the study period. Readers need this to understand who was included (STROBE items 5 and 6)."

Comment 2. First draft: "You should also do a cost-effectiveness analysis." Rewrite: delete it. A new analysis that was not planned is outside the study's question. If you think costs matter, raise it as a point for the discussion: "The authors might mention in the discussion that costs were not assessed."

Comment 3. First draft: "The conclusion is wrong." Rewrite: "Discussion, final paragraph: the conclusion says the program 'reduces' missed checks. As this is an observational study with possible unmeasured differences between groups, would 'was associated with fewer' be closer to what the data show? The limitations paragraph could also mention unmeasured confounding."

Each rewrite gives a location, a reason and a way forward.

Common pitfalls

  • Rewriting the paper in your own style. Point out unclear passages; don't impose your preferences.
  • Asking for unplanned analyses without a clear reason. They can make the final report worse [2].
  • Asking authors to cite your own work unless it is clearly needed, and then say so to the editor.
  • Vague comments such as "needs work" that give the authors nothing to act on.
  • A harsh or personal tone.
  • Reviewing outside your expertise without saying so. Tell the editor which parts you could not assess.
  • Sharing the manuscript or using its ideas before it is published.
  • Missing the deadline without telling the editor.

Quick self-check, in our own words

  • I have the expertise and time, and I have declared any competing interests.
  • I have kept the manuscript private and followed the journal's policy on software and AI tools.
  • My report opens with a short summary in my own words.
  • I have separated major points from minor ones and numbered them.
  • Each point says where, what, why and how to fix it.
  • I have checked the reporting against the right guideline.
  • I have not asked for unplanned analyses without good reason.
  • My tone is one I would be glad to receive.

Further reading

  • Kelly and colleagues give an overview of how peer review works, its benefits and its criticisms [1].
  • Provenzale and Stanley offer a step-by-step guide and template for reviewing a manuscript [6].
  • Related Stetleaf Academy articles: What peer reviewers look for in a manuscript.

References

  1. Kelly J, Sadeghieh T, Adeli K. Peer review in scientific publications: benefits, critiques, & a survival guide. EJIFCC. 2014;25(3):227-43. PMID: 27683470.
  2. Hopewell S, Collins GS, Boutron I, Yu LM, Cook J, Shanyinde M, et al. Impact of peer review on reports of randomised trials published in open peer review journals: retrospective before and after study. BMJ. 2014;349:g4145. doi:10.1136/bmj.g4145. PMID: 24986891.
  3. Cobo E, Cortés J, Ribera JM, Cardellach F, Selva-O'Callaghan A, Kostov B, et al. Effect of using reporting guidelines during peer review on quality of final manuscripts submitted to a biomedical journal: masked randomised trial. BMJ. 2011;343:d6783. doi:10.1136/bmj.d6783. PMID: 22108262.
  4. Committee on Publication Ethics. Ethical guidelines for peer reviewers [Internet]. COPE. Available from: https://publicationethics.org/guidance/guideline/ethical-guidelines-peer-reviewers
  5. International Committee of Medical Journal Editors. Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals [Internet]. ICMJE; [updated regularly]. Available from: https://www.icmje.org/recommendations/
  6. Provenzale JM, Stanley RJ. A systematic guide to reviewing a manuscript. AJR Am J Roentgenol. 2005;185(4):848-54. doi:10.2214/AJR.05.0782. PMID: 16177399.