Peer review protects the quality and integrity of what TQMJ publishes. As a reviewer you act as an independent expert who helps the editor decide whether a manuscript is scientifically sound, ethically conducted, clearly reported, and relevant to the journal's readership. Your task is not to rewrite the paper or to decide acceptance yourself; it is to evaluate the work honestly and to give the authors specific, actionable guidance for improvement.

A good review serves three parties at once: the editor, who needs a clear recommendation and reasoning; the authors, who need concrete direction; and future readers, who rely on the published record being trustworthy.

Accept an invitation only when you can review competently, impartially, and on time. Before agreeing, confirm each of the following:

•     Expertise. The topic falls within your area of knowledge so you can judge the methods and conclusions.

•     No conflict of interest. You have no personal, financial, academic, or institutional relationship with the authors that could bias your judgement (recent collaboration, shared institution, mentorship, competing work, or financial interest). If in doubt, declare it to the editor and let them decide.

•     Time. You can return a complete review within the agreed period (typically two to three weeks). If you cannot, decline promptly so the editor can invite another reviewer; if you need an extension, request it early.

•     Confidentiality. You accept that the manuscript is a privileged, unpublished document.

Confidentiality

Treat every manuscript as confidential. Do not share it, discuss it with others, or use any of its data, ideas, or results before publication. Do not upload the manuscript or any part of it to public websites or external tools, including generative-AI platforms, as this breaches author confidentiality. If you wish to involve a colleague or trainee in the review, obtain the editor's permission first and name them so their contribution can be acknowledged.

Conflicts of interest

Disclose anything that a reasonable observer might see as influencing your assessment. Declaring a minor connection does not automatically disqualify you; concealing one undermines trust in the process. When the conflict is substantial, decline and explain why.

Reviewers are expected to uphold the same ethical standards the journal applies to authors. In particular:

•     Be objective and constructive; criticise the work, never the authors. Hostile, dismissive, or personal remarks are not acceptable.

•     Raise any suspicion of misconduct — plagiarism, data fabrication or falsification, image manipulation, redundant or duplicate publication, or undisclosed ethical breaches — privately with the editor rather than confronting the authors directly.

•     Verify that human or animal research reports the relevant ethics-committee approval and informed consent. Flag any submission that lacks them.

•     Do not let nationality, religion, gender, seniority, or institution of the authors influence your assessment.

•     If you realise during the review that you have a conflict of interest, or that you are not competent to judge part of the work, tell the editor.

Read the manuscript twice: once quickly to grasp the overall question and findings, then again carefully with the checklist in hand. Assess the work section by section, and keep a numbered list of issues as you go — this becomes the backbone of your report.

4.1 Title and Abstract

•     Does the title accurately describe the study (design, population, setting) without overstating it?

•     Does the abstract match the body of the paper? Check that every figure quoted in the abstract reappears, unchanged, in the Results.

•     Are the main numbers internally consistent (percentages that sum correctly, counts that match the totals)?

4.2 Introduction

•     Is the rationale clear, and is the knowledge gap and aim explicitly stated?

•     Is the background supported by appropriate, current citations rather than broad unreferenced claims?

4.3 Methods

The Methods are the part you should scrutinise most closely, because conclusions are only as reliable as the design that produced them.

•     Study design: Is it appropriate for the research question and clearly described (cross-sectional, cohort, case–control, trial)?

•     Setting and period: Are the location and the data-collection dates stated and plausible? Check that the stated period is chronologically possible and consistent throughout the manuscript.

•     Participants: Are inclusion and exclusion criteria defined? Is the sampling method described, and is selection bias addressed?

•     Sample size: Is the chosen size justified by a power/sample-size calculation or by reference to a published estimate? A statement that the size is “sufficient” without calculation is not adequate.

•     Measurements and instruments: Are diagnostic kits, assays, or questionnaires identified by manufacturer and validated? For laboratory tests, are the reported sensitivity and specificity given, and is confirmation against a reference standard considered where relevant?

•     Statistics: Are the tests named and appropriate? Is the significance threshold stated? Where many comparisons are performed, is a correction (Bonferroni or Benjamini–Hochberg/FDR) applied? Is the software and version reported?

•     Ethics: Is approval from the relevant committee, with reference number, and informed consent documented?

4.4 Results

•     Do the results answer the stated aim, and are they presented without interpretation (interpretation belongs in the Discussion)?

•     Are tables and figures clear, correctly labelled, and non-redundant with the text?

•     Check the arithmetic: do row and column counts add to the totals, and do percentages in the text match the tables? Look for inconsistent notation (for example, “<6” versus “>6” used for the same group).

•     Are point estimates accompanied by 95% confidence intervals and exact p-values rather than “significant / not significant” alone?

•     For sparse tables, has the correct test (e.g., Fisher's exact) been used where expected cell counts are below five?

4.5 Discussion and Conclusion

•     Are the conclusions supported by the data, or do they overreach? A single borderline result should be interpreted cautiously, especially after multiple testing.

•     Is the work placed in context with comparable studies, and are agreements and disagreements described accurately and without internal contradiction?

•     Are limitations acknowledged honestly (design, sample, period, test performance, generalisability)?

•     Does the conclusion restate findings without introducing new, uncited claims?

4.6 References

•     Are references current, relevant, and formatted in a single consistent Vancouver style with correct journal abbreviations?

•     Verify that DOIs resolve and that dates are plausible. Be alert to references dated after the study period or that cannot be located — these warrant checking for accuracy or possible fabrication.

•     Watch for excessive or irrelevant self-citation.

TQMJ permits limited, declared use of AI tools by authors but holds them fully responsible for all content. As a reviewer, be alert to signs that text may have been machine-generated and inadequately checked, such as fluent but generic, filler phrasing sitting beside broken passages; confident statements without citations; contradictions a careful author would not leave in place; and references that look plausible but cannot be verified. Where you see such signs, note them factually and recommend that the editor ask the authors to declare any AI use, take responsibility for accuracy, and provide raw data and original statistical output. Never paste the manuscript into an AI tool yourself — that breaches confidentiality.

Your report has two separate parts, and it is important to keep them distinct:

•     Confidential comments to the editor. Your overall judgement, your recommendation, and any sensitive concerns (suspected misconduct, conflicts). Do not put your recommendation here in a way the authors will see.

•     Comments to the authors. Specific, numbered, and actionable. This is what the authors will use to revise.

Structure the comments to authors as follows:

1.   A brief summary (two to three sentences) showing you understood the study and stating its main contribution. This reassures authors that the review is informed.

2.   Major points — issues that affect the validity of the conclusions (design flaws, statistical errors, unsupported claims, missing ethics approval). Number them.

3.   Minor points — issues that improve clarity and presentation (wording, table formatting, typographical and grammatical errors, reference style).

Be specific. “The statistics are weak” helps no one; “Nine comparisons are made without correction; apply an FDR correction and re-interpret the rotavirus–sex association accordingly” does. Quote the page, table, or line where you can. Phrase criticism professionally and assume good faith.

Conclude with one clear recommendation to the editor. Match your recommendation to the substance of your comments — a recommendation of “minor revision” alongside several major concerns confuses the editor and the authors.

Recommendation	When to use it
Accept	The manuscript is sound and ready, or needs only trivial edits. Rare on first review.
Minor revision	The work is fundamentally sound; the remaining issues are limited and presentational and do not require re-review of the science.
Major revision	The study has merit and usable data, but substantial issues (methodological, statistical, or interpretive) must be resolved and the revision re-reviewed before a decision.
Reject	The study has fatal flaws that revision cannot fix (invalid design, unsound data, unethical conduct), or it is out of scope. Explain the reasoning respectfully.

When you recommend revision, state clearly which points are essential conditions for acceptance and whether you are willing to re-review the revised version.

Use this list as a final pass before submitting your report. It is a prompt, not a substitute for narrative comments.

Scope and presentation

☐  Topic is within the journal's scope and of interest to its readers.

☐  Title and abstract accurately reflect the study and match the body.

Methods

☐  Design is appropriate and clearly described.

☐  Setting and study dates are stated, plausible, and consistent.

☐  Inclusion/exclusion criteria and sampling are defined.

☐  Sample size is justified (power calculation or cited estimate).

☐  Instruments/assays are identified and validated; test performance reported.

☐  Statistical tests are named, appropriate, and corrected for multiple comparisons where needed.

☐  Ethics approval (with number) and informed consent are documented.

Results and discussion

☐  Numbers are internally consistent (counts, totals, percentages, notation).

☐  Estimates include 95% confidence intervals and exact p-values.

☐  Tables and figures are clear, correct, and non-redundant.

☐  Conclusions are supported by the data and not overstated.

☐  Comparisons with other studies are accurate and free of contradiction.

☐  Limitations are acknowledged honestly.

Integrity and references

☐  No signs of plagiarism, data or image manipulation, or undisclosed AI use.

☐  References are consistent Vancouver style; DOIs resolve; dates are plausible.

☐  No conflict of interest on the reviewer's part; confidentiality maintained.

You may copy the structure below into your report.

A. Confidential comments to the Editor

Manuscript ID / Title
Reviewer name
Date
Overall recommendation	Accept / Minor revision / Major revision / Reject
Conflict of interest	None / Declared (specify)
Willing to re-review	Yes / No
Confidential notes to editor