All research articles, AI and prediction model studies, intervention evaluations, reviews, implementation reports, software and tool papers, data descriptors, short communications and study protocols submitted to JMIDH undergo double-blind peer review: authors and reviewers do not know each other’s identities. Viewpoints, editorials and letters are assessed by the editors and marked as not externally peer reviewed.

Stages and timelines

Technical check (3 to 5 working days). The editorial office checks completeness, anonymisation (including code links), declarations, the reporting checklist, ethics documents and the Turnitin similarity report.

Editorial screening (within 10 working days). The Editor-in-Chief or an Associate Editor assesses scope, originality, methodological soundness, clinical or health relevance and ethics. Manuscripts may be declined without external review, with reasons given.

Reviewer invitation (within 7 days). The handling editor invites at least two independent external reviewers. For AI and data science papers, at least one reviewer has technical or methodological expertise and at least one has clinical or health-domain expertise.

Peer review (21 days). Reviewers assess the manuscript and, where possible, inspect the anonymised code and documentation. Software papers are checked for installability and documentation.

Methods and statistical review (as needed). Requested for prediction models, trials, meta-analyses and complex analyses.

First decision (target 6 to 8 weeks from submission). The handling editor recommends a decision and the Editor-in-Chief confirms it.

Revision. Authors revise within 15 days for minor revisions and 45 days for major revisions and provide a point-by-point response.

Reproducibility check and final decision. Before acceptance, the Reproducibility Editor confirms that code and data statements are accurate and links are working. Revised manuscripts normally return to the original reviewers.

Production (3 to 4 weeks after acceptance). Copy-editing, typesetting, proofs, DOI registration and online publication.

Anonymity

Authors submit a separate title page and an anonymised manuscript. Code repositories must be anonymised for review. Reviewers must not attempt to identify authors, including by searching for preprints, conference papers or repositories. Reviewer identities are never revealed to authors.

Reviewer selection

Reviewers are chosen for relevant expertise and verified through institutional email, ORCID or publication profiles. They must not work at the same institution as any author, must not have co-authored with them in the last three years, and must not have a commercial interest in a competing product. Authors may suggest up to three reviewers and request exclusions with reasons, but suggested reviewers are never the only reviewers. A further reviewer may be invited if reports conflict.

Decisions

Possible decisions are accept, minor revision, major revision, reject with encouragement to resubmit, or reject. The editor’s decision is based on reviewer reports, scientific quality, reproducibility, safety and ethics, not on a count of reviewer votes. Authors receive all anonymised reviewer comments. Acceptance cannot be guaranteed in advance, and because JMIDH charges no fees, no payment is ever involved in the process.

Manuscripts by editors and board members

Submissions from the Editor-in-Chief, editors, board members or publisher staff are handled entirely by another editor with no conflict. The author-editor has no access to the review process. The same double-blind process applies, and the published article discloses the author’s editorial role.

Confidentiality and AI

Manuscripts, code and data under review are confidential. Editors and reviewers must not upload them, or their review reports, to generative AI tools or other external services.

Appeals

Authors may appeal a decision once, following the Appeals Process.