Thank you for considering the Journal of Medical Informatics and Digital Health (JMIDH). Please read these instructions carefully; manuscripts that do not follow them will be returned before peer review. JMIDH follows the ICMJE Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals. JMIDH does not charge any fees for submission, processing or publication.

1. Before you submit

•     Check that your work fits the Aims & Scope. Studies must involve health data, a health problem or a healthcare setting.

•     The manuscript must be original, not published elsewhere (except as a preprint, thesis or conference abstract) and not under consideration by another journal. Work previously published in full conference proceedings (for example, IEEE or ACM proceedings) may be considered only if substantially extended — normally at least 30% new content — and the earlier paper is cited and disclosed.

•     All authors meet the authorship criteria and approve the submission.

•     Ethics approval or a formal exemption is in place for studies using patient data, and trials of digital interventions are registered.

•     Code, models and data will be made available as described in the Code, Model & Data Availability policy, or restrictions are explained.

2. Article types

Word counts exclude the title page, abstract, references, tables, figure legends and declarations. Editors may allow longer manuscripts when justified.

Original Research. Quantitative, qualitative or mixed-methods studies of health information systems, digital interventions or health data. Up to 4,000 words; structured abstract up to 300 words; up to 50 references; up to 6 tables and figures.

AI and Prediction Model Study. Development, validation, updating or impact assessment of prediction models, machine learning, imaging AI or language models. Up to 4,500 words; structured abstract up to 300 words; up to 50 references; up to 8 tables and figures, plus supplementary material. External validation studies are particularly welcome.

Evaluation of a Digital Intervention. Randomised or non-randomised evaluation of an app, telemedicine service, decision support system or other digital intervention. Up to 4,000 words; structured abstract up to 300 words; up to 50 references; up to 6 tables and figures.

Systematic Review, Meta-analysis or Scoping Review. Following PRISMA 2020 or PRISMA-ScR; prospective registration is strongly encouraged. Up to 5,500 words; structured abstract up to 300 words; no reference limit; up to 8 tables and figures.

Narrative Review. A critical, current review with a documented search strategy. Up to 5,000 words; unstructured abstract up to 250 words; up to 100 references.

Implementation Report. Describes the deployment of a digital health system in a real setting, with data on adoption, use, outcomes, costs and lessons learned, following iCHECK-DH. Up to 3,500 words; structured abstract up to 250 words; up to 30 references; up to 5 tables and figures.

Software and Tool Paper. Describes open-source software, a library, algorithm implementation or tool of value to health informatics, with evidence of testing or use. The code must be publicly available under an open licence with an archived version that has a DOI. Up to 3,000 words; unstructured abstract up to 200 words; up to 30 references.

Data Descriptor. Describes a health dataset that is openly available or available under controlled access, including collection methods, de-identification, quality checks and access conditions. The dataset must be deposited in a recognised repository. Up to 3,000 words; unstructured abstract up to 200 words; up to 30 references.

Short Communication. Concise report of focused or preliminary findings. Up to 2,000 words; structured abstract up to 200 words; up to 15 references; up to 2 tables and figures.

Study Protocol. Protocols of registered, ethically approved studies not yet completed. Up to 4,000 words; structured abstract up to 300 words.

Viewpoint. Evidence-based opinion on policy, regulation, ethics or practice in digital health. Up to 2,000 words; no abstract; up to 20 references.

Letter to the Editor. Comment on an article published in JMIDH within the past six months, or a brief observation. Up to 800 words; up to 5 references.

Editorial. By invitation of the Editor-in-Chief. Up to 1,200 words.

3. Files required at submission

1.    Cover letter explaining the contribution and its relevance to health or healthcare, and disclosing any related conference papers or preprints.

2.    Title page (not sent to reviewers) — see section 4.

3.    Anonymised main manuscript. Remove author names, affiliations, hospital and institution names, ethics committee names, acknowledgements and funding. Replace such details with ‘[blinded for review]’. Remove names from file properties.

4.    Anonymised code and data links for review. Repositories such as GitHub reveal author identities. For review, provide an anonymised copy (for example through Anonymous GitHub or a private link shared through the editorial office) or upload the code as a supplementary file with identifying details removed. Public links are added after acceptance.

5.    Tables in editable form, and figures as separate high-resolution files. Screenshots of software must not show real patient data.

6.    Reporting guideline checklist with page numbers (for example TRIPOD+AI, CONSORT-EHEALTH, CLAIM or STROBE).

7.    Ethics approval letter or exemption letter.

8.    ICMJE disclosure forms for all authors.

9.    Supplementary material such as model details, hyperparameters, prompts, extended results and survey instruments, anonymised.

4. Title page

The title page must include a concise, informative title stating the design (for example ‘…: development and external validation of a machine learning model’, or ‘…: a randomised controlled trial’); a running title of no more than 50 characters; the full names, highest degrees, departments, institutions, cities, countries, email addresses and ORCID iDs of all authors (an ORCID iD is mandatory for the corresponding author); the corresponding author’s postal address; the word count and numbers of tables and figures; and all declarations listed in section 7.

5. Manuscript structure for research articles

•     Structured abstract with Background, Objective, Methods, Results and Conclusions, including the trial registration number where applicable.

•     Keywords: 4 to 6, preferably MeSH terms.

•     Introduction: the problem, current evidence and the gap addressed.

•     Methods: setting and data sources with dates; participants and eligibility; for AI studies, the outcome and its definition, predictors, handling of missing data, data splitting (with patient-level separation to avoid leakage), model development, hyperparameter tuning, validation strategy (internal and preferably external), performance measures (discrimination, calibration and clinical utility), and fairness assessment across subgroups; for digital interventions, a description of the intervention and its version, delivery, usage and adherence; statistical analysis; ethics and consent.

•     Results: following the reporting guideline, with confidence intervals for performance measures and effect estimates.

•     Discussion: principal findings, comparison with previous work, strengths and limitations (including generalisability, bias and data quality), implications for practice, policy and future research.

•     Conclusions supported by the data, without overstating readiness for clinical use.

Studies of large language models must also report the model name, version and access date, how it was accessed (web interface or API), all prompts (in full in supplementary material), settings such as temperature, the number of runs, how outputs were evaluated and by whom, and how patient confidentiality was protected.

6. Style

•     English, British or American, used consistently. Non-native English is not a reason for rejection if the science is sound.

•     Word format, 12-point font, double spacing, with line and page numbers.

•     Define abbreviations at first use. Use standard names for coding systems and standards (for example SNOMED CT, HL7 FHIR R4).

•     Name software, libraries and versions (for example Python 3.12, scikit-learn 1.5) and cite them where a citation is available.

•     Commercial products may be named with the manufacturer and country at first mention; avoid promotional language.

References

Use the Vancouver (ICMJE/NLM) style, numbered in order of citation, with DOIs where available. Cite datasets, software and code repositories in the reference list with their persistent identifiers. Preprints may be cited and must be labelled as preprints. Authors are responsible for the accuracy of every reference; references that cannot be verified will be queried.

7. Declarations

Each article must include the following statements. Ethics approval: the committee name, reference number and date, or the reason approval was not required, with the exemption letter. Consent: how consent was obtained, or that the ethics committee waived it and why. Trial registration: registry and number. Author contributions: CRediT roles for every author. Funding: funders and grant numbers and the role of the funder, or a statement that there was no specific funding. Competing interests: including relationships with technology companies, vendors and start-ups, or ‘The authors declare no competing interests’. Data availability and Code availability: see P19. Use of generative AI: the tool, version and purpose, or ‘No generative AI tools were used in preparing this manuscript’. Acknowledgements: contributors who are not authors, with their consent.