1. The editor’s decision-making responsibility
Editorial decisions are reasoned judgments about the manuscript, the evidence presented, the journal’s remit and the standards applicable to the work. Peer reviewers provide independent advice; they do not vote on publication, and the handling editor should not outsource the final judgment to the numerical balance of recommendations.
An editor should be able to identify which findings in the reviews materially affect the decision, which issues can be resolved by revision, and which problems prevent the work from meeting the journal’s threshold. Consistency matters, but identical outcomes are not required for manuscripts with different evidence and risks.
Apply publisher-level mandatory policies alongside documented journal-specific requirements. When an issue lies outside the editor’s expertise, obtain targeted advice instead of filling the gap with assumptions.
2. Read and evaluate reports before counting recommendations
Read the substance of each review independently of its accept, revise or reject recommendation. Assess whether criticisms are specific, evidence-based, within the reviewer’s expertise and relevant to the claims made in the manuscript. A forceful tone does not make a weak argument stronger.
Distinguish factual errors, methodological limitations, interpretive disagreements, requests for additional experiments and matters of presentation. The first three may affect the validity of the work; the latter two require proportionality and a clear link to the manuscript’s stated contribution.
Where a report is superficial, internally contradictory, inappropriate or dominated by irrelevant citation requests, the editor should not pass it to authors uncritically. Seek clarification, discount unsuitable portions or commission targeted additional review if a material assessment gap remains.
3. Synthesize evidence across conflicting reviews
Disagreement among reviewers is common, especially for interdisciplinary or innovative research. Determine whether they disagree about facts, the importance of a limitation, methodological interpretation or the journal’s publication threshold. These are different problems and require different editorial responses.
Build an issue-by-issue synthesis rather than averaging recommendations. A single well-supported methodological objection can be consequential even when other reviewers are enthusiastic; similarly, a strongly negative recommendation may carry little weight if its reasons are unsupported or outside scope.
Additional review is warranted when it can answer a clearly identified question that the existing reports cannot resolve. Do not seek further reviewers merely to obtain a majority or a preferred outcome.
4. Define material flaws and revision feasibility
Ask whether the central claims are supported by the design, data and analyses available. Separate a fundamental defect that cannot realistically be corrected from a remediable problem in reporting, analysis, framing or interpretation. A limitation may be acceptable when it is accurately acknowledged and the conclusions are proportionate.
Requests for additional experiments or datasets should be justified by what is necessary to support the paper’s principal claims, not by an idealized expansion of the project. An editor should be prepared to explain why a requested change is essential, optional or beyond the reasonable scope of revision.
Where a revised analysis or narrower claim could address the concern, communicate that pathway precisely. Where the work cannot support its central conclusion even after plausible revision, repeated review cycles impose unnecessary costs on authors and reviewers.
5. Choose among editorial outcomes
An initial or post-review decision may be to decline the manuscript, invite major revision, invite minor revision, proceed toward acceptance subject to defined checks, or take another route permitted by the journal’s workflow. The label should reflect the actual outstanding work and should not be used to soften or obscure the substance of the decision.
Major revision is appropriate when important scientific or methodological questions remain open and the editor needs to evaluate a substantive response. Minor revision should be reserved for limited issues unlikely to alter the principal scientific assessment. Acceptance should follow only when the manuscript satisfies the journal’s applicable requirements.
A revision invitation is not a promise of acceptance. Conversely, a request for changes should not imply an unlimited opportunity to redesign a study. State the status of the manuscript, the principal issues and what the editor expects to evaluate next.
6. Make desk rejection and post-review rejection intelligible
Desk rejection and rejection after review occur at different stages and should be communicated accordingly. A desk decision may rest on clear scope mismatch, unmet journal criteria or an evident problem that makes productive external review unlikely. A post-review rejection should reflect the editor’s assessment of the manuscript and relevant review evidence.
Give the authors the main reasons in professional, manuscript-specific language. Avoid generic phrases when a concise explanation is possible, but do not turn a decision letter into an exhaustive second review or disclose confidential internal deliberations.
Do not make misconduct allegations in a rejection letter merely because a reviewer raised a concern. Handle credible integrity issues through the relevant confidential process and separate allegations from established editorial reasons.
7. Write a coherent and actionable decision letter
Begin with the decision and a short editorial synthesis. Identify the issues that drive the outcome, then distinguish essential changes from optional improvements. Authors should not have to infer which of several contradictory reviewer demands they are expected to follow.
When transmitting reviewer reports, remove inappropriate language and handle confidential editor-only comments according to journal policy. If a reviewer request is unsupported, irrelevant or inconsistent with the journal’s standards, the editor should clarify that it is not a condition of revision.
Use precise and respectful language. A clear letter explains what the manuscript needs, why those changes matter and what the next stage will be. Avoid promising acceptance or requiring work that the editor has not judged necessary.
8. Prioritize reviewer requests and resolve contradictions
Create a hierarchy of concerns: issues affecting validity, participant protection or integrity first; then major interpretive and reporting problems; finally presentation and minor corrections. This hierarchy helps authors allocate effort and reduces the chance of repeated rounds driven by shifting editorial expectations.
When reviewers offer mutually incompatible requests, the editor should decide which request is necessary, whether both are optional, or whether a focused clarification is needed. Simply forwarding contradictions transfers the editor’s responsibility to the authors.
Requests for new citations should be evaluated for direct scholarly relevance. Citation demands intended to benefit a reviewer, editor or journal should not be treated as conditions of publication.
9. Manage revision cycles proportionately
Set revision deadlines that reflect the work requested, and explain whether the revised manuscript is likely to return to reviewers. An extension may be reasonable for substantive work, but the journal should maintain transparent and consistent workflow expectations.
On resubmission, assess whether the authors have addressed the decision letter and whether changes create new questions. The response letter is an aid to evaluation, not a substitute for checking the revised manuscript and relevant analyses.
Avoid endless cycles in which each round introduces unrelated new demands. New requirements should generally arise only when revisions expose a material issue or a previously overlooked concern is genuinely consequential.
10. Decide when to return a revision to reviewers
Return a revision to reviewers when their expertise is needed to evaluate whether a central scientific or methodological concern has been resolved. Routine formatting changes, straightforward clarifications and minor editorial corrections often can be checked without a new full review.
Where possible, invite the original reviewers whose substantive concerns are being addressed, subject to availability and continuing independence. If an original reviewer is unavailable, decide whether a new specialist is necessary or whether the editor can assess the change.
Provide reviewers with the relevant revised manuscript, response and prior comments through the journal’s approved system. Explain the scope of the reassessment so that the new round does not become an unrelated evaluation of the entire project.
11. Handle statistical and specialist uncertainty
Editors should recognize when an issue requires expertise they do not personally possess. A methodological disagreement, unusual analysis or specialized technical claim may warrant consultation with a qualified specialist reviewer or statistical editor.
Frame the consultation around a defined question: whether the analysis supports a particular claim, whether a reporting omission prevents assessment, or whether a proposed revision resolves a specific weakness. Vague requests for another general opinion can add delay without reducing uncertainty.
Specialist advice informs but does not replace editorial judgment. The handling editor should synthesize technical findings with the journal’s remit, the paper’s contribution and other relevant review evidence.
12. Integrate ethics, integrity and confidentiality into decisions
Ethical approval, consent, trial registration, data integrity, image concerns, duplicate publication and other policy issues may arise during review. Treat a missing statement differently from a credible substantive concern; some matters require straightforward clarification, while others require a confidential integrity pathway.
Do not ask peer reviewers to conduct investigations or independently contact institutions, participants or authors. Preserve relevant material and escalate credible concerns through authorized editorial procedures. Publication decisions should not pre-empt a fair assessment of unresolved allegations.
Protect reviewer identities and unpublished manuscript information according to the journal’s review model. Do not place confidential review content into unapproved external or generative-AI services when drafting or evaluating decisions.
13. Preserve editorial independence and manage conflicts
The handling editor should consider whether relationships with authors, reviewers, institutions or the subject matter create an actual or reasonably perceived competing interest. Where the journal’s rules require recusal, the manuscript should be reassigned before substantive decisions are made.
Editors should not allow commercial pressure, institutional influence, personal rivalry, citation incentives or anticipated publicity to determine a manuscript’s scientific outcome. Publisher support and journal governance should preserve the independence of legitimate editorial assessment.
An editor who discovers a material conflict late in the process should disclose it promptly and seek reassignment or other management under policy. The editorial record should explain consequential changes without unnecessarily disclosing private information.
14. Document reasoning and maintain a decision audit trail
Record the decision, principal evidence, material reviewer concerns, how disagreements were resolved and the reasons for unusual procedural choices. Documentation should enable an authorized colleague to understand the editorial path without relying on the handling editor’s memory.
Do not reduce the record to a vote tally. A defensible decision note identifies the scientific and policy considerations that mattered and distinguishes unresolved uncertainty from a proven deficiency. Notes should remain factual, professional and proportionate.
Preserve decision letters, reviews, revisions, conflict disclosures and escalation records in the journal’s approved systems. Access should follow confidentiality and retention rules rather than convenience.
15. Respond to appeals and complaints without defensiveness
Authors may challenge an editorial decision or raise a procedural complaint. Explain the journal’s published route and assess whether the appeal identifies a material factual error, procedural irregularity, overlooked evidence or another basis recognized by policy.
An appeal is not automatically another review round. Where appropriate, a different editor or authorized body should assess the issue to preserve independence, particularly when the original decision-maker has a conflict or is the subject of the complaint.
Communicate the outcome and the reasoning permitted by confidentiality requirements. Do not penalize authors merely for using an established appeal or complaint process.
16. Apply a final pre-acceptance verification
Before acceptance, confirm that the substantive editorial conditions have been met and that required declarations, ethical statements, data or materials information and other journal-specific requirements are appropriately addressed. Production-stage formatting should not substitute for scientific or policy checks.
Check whether final claims remain aligned with the evidence after revision, whether key figures and tables correspond to the accepted text, and whether material issues identified during review have a documented resolution.
Acceptance closes the scientific evaluation stage but does not eliminate later responsibilities. Any concern discovered after acceptance should be handled promptly under the appropriate editorial or research-integrity procedure.
17. Communicate consistently with authors and reviewers
Timely, respectful communication is part of a fair editorial process. Authors should understand the decision and the required next action; reviewers should receive appropriate acknowledgment and should not be exposed to unnecessary confidential details about internal deliberations.
When substantial delays occur, communicate what can be shared without making promises about outcomes. Do not describe reviewers as the final decision-makers or attribute editorial choices to them in a way that evades editorial accountability.
Where the journal provides reviewer feedback or decision notifications, follow the applicable review model and confidentiality rules. Communications should support trust without exposing protected identities or correspondence.
18. A repeatable editorial decision framework
At each decision point, identify the manuscript’s central contribution, the evidence that supports or limits its claims, the material issues raised by reviewers and any policy requirements that must be satisfied. Ask which issues are resolvable, which require specialist advice and which fundamentally prevent progression.
Select an outcome that matches the actual outstanding work. Write a decision letter that prioritizes essential concerns, resolves contradictory requests and explains the next stage. Keep a factual record of the reasoning and any exceptional handling.
The goal is not unanimity among reviewers or an arbitrary number of revision rounds. It is a proportionate, independent, evidence-based and documented editorial decision that can be explained to authors and defended through the journal’s governance process.