Aims
The Journal of Reproducible Health Sciences publishes health research whose results are reproducible as far as the data allow. Health evidence changes what clinicians do, what programmes are funded and what regulators permit, and it is very often built on records that cannot simply be posted online. That combination (high stakes, low inspectability) is what this journal is built to change. Every submission ships the code and data behind its results, or a privacy-preserving surrogate where the records cannot be shared, and a named human editor reads and signs every decision.
Scope
Scope is empirical health research with a verifiable evidence substrate — epidemiology, public and global health, clinical and rehabilitation sciences, health data science and biostatistics, and environmental and occupational health. Privacy-bound data are in scope through the declared data-restriction routes. Study protocols and statistical analysis plans are in scope in every section, and are assessed as plans for work not yet done.
The clause about privacy-bound data is deliberate. A confidentiality constraint is a fact about the data, not a verdict on the work: declared, with its access route stated, it keeps a paper inside scope and sends it down the restricted-data route rather than out of the journal. What scope does require is an examinable evidence substrate, data, an analysis, a model or code that someone other than the authors could inspect.
Sections
The journal publishes as one title, with one editorial board and one shared citation record, in six named sections:
- Epidemiology. Observational designs, causal inference from observational data, surveillance, and register-based research.
- Public & Global Health. Population interventions, policy evaluation, health systems and services, and research in low-resource settings.
- Clinical & Rehabilitation Sciences. Trials, physiotherapy and rehabilitation research, sleep and behavioural health, clinical measurement and diagnostic accuracy.
- Health Data Science & Biostatistics. Electronic health records and routinely collected data, prediction models, wearables and digital health, and statistical methods applied to health questions.
- Environmental & Occupational Health. Exposures, workplace health, and climate and health.
- Study Protocols & Statistical Analysis Plans. The plan of a study, published before its results exist, so that the finished work can be held to it. Described in full below.
Study protocols and statistical analysis plans
The journal publishes the protocol of a planned study, and the statistical analysis plan that belongs to it, as an article in its own right. It is citable, it carries a DOI, and it appears before the results exist. A protocol may be submitted in any of the five sections.
The reason is the same one the journal is built on. A plan written down and published before the data arrive is what makes the later results paper checkable against something other than itself. It fixes the primary outcome, the analysis model and the handling of missing data at a point where nobody yet knows which answer they would prefer.
What review asks of a protocol is whether it is specific enough to be held to later. That means a stated primary outcome with its measurement, the analysis model and its assumptions, the handling of missing data and of deviations, the sample size and what it was calculated to detect, and any stopping or interim rules. Trial protocols are expected to follow SPIRIT 2025 and to ship the completed checklist, in the same way other designs ship theirs. An interventional study still needs its prospective registration before it can be considered.
A protocol carries no reproducibility badge, because there are no results to reproduce. Code that already exists travels with it all the same, most often the simulation behind the power calculation and the randomisation code.
This is not a registered report. Publishing a protocol here says nothing about the later results paper, which is submitted, reviewed and decided on its own. We make no commitment in advance to publish it.
Research on people
This journal publishes research on human health: human subjects, populations and health systems. Research on people is admissible only with the approvals it required, a named ethics committee and its reference number, documented consent or a documented reason none was needed, and handling of individual-level data that respects the law it was collected under (see publication ethics).
Animal-model studies are out of scope.
Interventional studies and prospective registration
Trials and other interventional studies are in scope with prospective registration. As a condition of consideration, the study must be registered in a WHO ICTRP primary registry or ClinicalTrials.gov at or before the time of first patient consent for enrolment, and the manuscript must state the registration identifier, the registry and the registration date.
Review checks that a registration is stated, that it is prospective against the manuscript's own dates, and that the registered primary outcome and the reported one are the same study, all from the manuscript itself. The registry entry is then checked by the editors against the registry before a decision is signed.
An interventional study that was not prospectively registered cannot be considered.
Reporting guidelines
Submissions are expected to follow the EQUATOR-matched reporting guideline for their design, and to ship the completed checklist with the manuscript:
- CONSORT 2025 — randomised trials.
- STROBE — observational studies, with RECORD for studies using routinely collected health data.
- TARGET — observational studies emulating a target trial, alongside STROBE.
- PRISMA 2020 — systematic reviews and meta-analyses.
- TRIPOD+AI — development and validation of clinical prediction models.
- STARD — diagnostic-accuracy studies.
- CHEERS 2022 — economic evaluations.
Where no listed guideline fits a design, say which one comes closest and where the study departs from it.
Systematic reviews and meta-analyses
Actively welcome, and reviewed on equal terms. A synthesis is among the most reproducible genres in health research when it travels with its search strategy, its extraction table and its analysis code, and among the least when it does not. We ask for all three, and we rerun the analysis.
What we cannot review well yet
Qualitative health research is out of scope for now. The review roster this journal runs is quantitative, and it would judge an interview study by criteria that do not fit it. A qualitative-methods reviewer profile comes first. Mixed-methods work with a quantitative core is in scope, and its qualitative strand is read for transparency of method and material.
Case reports are out of scope. A single clinical narrative carries no substrate this journal's verification can act on.
Privacy-bound data, and what the badge claims
Every contribution carries a reproducibility package (an executable source, the analysis code, the data or a surrogate, a pinned environment and its determinism information) and where individual-level records cannot be released, the restriction is declared with the route by which a reader could obtain access, and an anonymised or synthetic surrogate takes the data's place wherever one can be built.
The badge then reports what the check produced. ✓ Reproduced means the deposited code, over the deposited data, regenerated the results as printed. ◑ Partially reproduced means some of them did. Privacy-limited means the original records could not be released. It is a separate state and does not rank below the other two. In every case the article's reproducibility panel states in words what was verified and what was not. See data & code availability for the package and the restricted-data routes.
How claims are read in review
Health papers are read against their design as well as their numbers. Review asks whether the sample and the analysis can support the conclusion drawn, whether the design licenses the causal language used, whether the abstract's claim is the one the results establish, and whether the reporting covers multiplicity, missing data and the ways selection could have produced the finding. None of that is about how interesting the topic is.
Out of scope
- Work with no examinable evidence at all: no data, no analysis, no model, no code, and no willingness to document how the results were produced. A confidentiality restriction is not a reason for exclusion. It has a route.
- Research on people conducted without the required ethical approval, or reporting individual-level data the authors were not permitted to analyse or publish.
- Animal-model studies, qualitative-only studies and case reports, for the reasons given above.
- Work with no health question behind it, and topics outside the health sciences.
- Advocacy without evidence, and evidence assembled to reach a commissioned conclusion.
Soundness, not newsworthiness
We judge a submission on scientific soundness and reproducibility, not on novelty or how newsworthy the finding seems. A rigorous, correctly reported, reproducible study showing that an intervention did not work is in scope and publishable, and it is the kind of paper the literature is most short of, because it is the kind that is hardest to place. This is not "accept everything": the bar is rigour plus reproducibility. Replications, reanalyses and null findings are welcomed deliberately.
How submissions are assessed, including the disclosed use of AI and the named human editor who signs every decision, is described in the peer-review policy. What a submission must contain is on the author guidelines page.