Nature Medicine sets a high bar for clinical research transparency
Nature Medicine’s clinical research policy reads less like a submission checklist than a compliance framework for modern trial science. The journal will consider all types of clinical research, but the path to review is narrow: prospective registration before the first participant is enrolled, a clearly stated trial number in the abstract and methods, and close alignment between the reported analysis and the prespecified protocol and statistical analysis plan. Those requirements matter because the credibility of a clinical trial lives or dies on whether the final manuscript can be traced back to a locked design, not a retrospective reconstruction.
Registration is not a formality here. The journal requires trials to be entered in an acceptable registry before enrollment begins, with acceptable registries including WHO ICTRP primary registers and ClinicalTrials.gov. The policy also restates the ICMJE definition of a clinical trial: any study that prospectively assigns people or groups to an intervention, with or without a concurrent comparison group, to examine a health-related intervention and a health outcome. In practice, that definition sweeps in drugs, surgery, devices, behavioral treatments, educational programs, dietary interventions, quality improvement interventions, and process-of-care changes. By drawing that boundary explicitly, the journal makes clear that intervention studies cannot evade trial standards simply because they do not involve a conventional drug product.
Primary results from nonregistered trials will not be considered. That is a hard stop.
Protocol fidelity and prespecified analysis sit at the center of review
The submission package for a clinical trial does not end with the manuscript itself. Investigators must provide the most recent clinical trial protocol in English, signed and dated, along with any amendments and a summary of why those changes were made. If a separate statistical analysis plan exists, that document must arrive too, and it must contain enough detail for another analyst to reproduce the work in a similar dataset. When the original protocol or SAP is in another language, the journal expects machine-readable versions of those source documents.
The policy gives editors and reviewers a clear lens for judging the analysis: the primary outcome must be handled exactly as specified in the final protocol and SAP, except in unusual circumstances. Deviations from the registry or protocol require justification in the Methods section. This is not bureaucratic fussiness; it is the mechanism that protects trials from selective emphasis, outcome switching, and post hoc narrative drift. A trial that changes its analytic rules after seeing the data can generate a polished paper and a weak inference.
Interim analyses receive similar scrutiny. They must be prespecified in the protocol. Unplanned looks at ongoing data may be considered, but only case by case, and any post hoc or exploratory analysis must be labeled as such in the abstract and manuscript text. That distinction matters because early peeks at the data can distort interpretation, especially when safety signals, efficacy trends, or subgroup patterns tempt investigators to overread unstable findings.
Reporting standards now extend well beyond randomized trials
Nature Medicine aligns clinical reporting with the major international frameworks. Randomized trials must follow CONSORT 2025, cluster randomized controlled trials need the cluster extension, and nonrandomized trials are encouraged to use CONSORT principles as well. Observational cohort, case-control, and cross-sectional studies must follow STROBE. Systematic reviews and meta-analyses belong under PRISMA. Biomarker studies tied to clinical outcomes must use STARD or an appropriate extension, with oncology biomarker studies expected to follow REMARK. Prediction model papers fall under TRIPOD, or TRIPOD-AI when machine learning or regression methods are used. Observational work that tries to estimate causal effects by emulating a target trial must follow TARGET.
These are more than citation preferences. Each framework is a check against common failure modes in clinical literature: incomplete flow reporting, opaque inclusion criteria, missing adverse event data, and endpoints introduced after the fact. For trial reports, the journal specifically recommends a CONSORT-style flow diagram, a baseline characteristics table, adverse events reported in the main text, and exact dates for first enrollment, last enrollment, and data lock. Even the requirement for exact dates reflects a central truth of clinical research: without a clear temporal sequence, readers cannot judge whether the evidence was generated under stable conditions or during a moving target of evolving practice.
Data sharing, conflicts, and sponsor role are part of the evidentiary record
The editorial policy treats data availability as a core component of trust. Every original research manuscript needs a Data Availability Statement describing what data exist, how to access them, when they can be obtained, and what restrictions apply. If custom code drives the analysis, a separate Code Availability Statement is required. Supporting source data and code must be available to editors and peer reviewers when requested. For trials that began enrolling on or after 1 January 2019, the registry record must include a data sharing plan, and any later change in that plan must be reflected both in the statement submitted with the manuscript and in the registry itself.
Transparency also extends to influence. Authors must disclose competing financial and non-financial interests in a dedicated section, and the corresponding author carries responsibility for submitting the statement on behalf of the full author group. Funding must be declared separately, along with the sponsor’s role at any stage of the work, including manuscript preparation. If the sponsor had no role, that must also be stated explicitly.
The policy’s structure leaves little room for ambiguity about editorial expectations. Researchers are expected to show their work, not just report their conclusions.
Why these rules matter at the bedside and in the trial office
Clinical research that reaches practice without transparent registration, protocol discipline, and traceable data handling can mislead clinicians just as easily as it can inform them. A dietary intervention, a process-of-care change, or a device study can influence patient outcomes as powerfully as a drug trial, which is why the journal applies the same evidentiary logic across intervention types. Prespecified endpoints, documented amendments, and a visible audit trail reduce the chance that an apparently positive result is really a product of selective analysis or incomplete reporting.
For investigators, the policy raises the threshold for submission. For readers, it raises confidence in the manuscript when the paper survives that threshold. The practical consequence is straightforward: a clinical trial that cannot be reconstructed from its protocol, registry entry, and analysis plan will not fit this journal’s standards, even if the narrative looks polished on the page. The next layer of scrutiny will come from the protocol, the SAP, the registry record, and the data-sharing statement, not from rhetorical polish in the Discussion section.
The policy also signals what the journal expects from modern clinical science: not just statistically significant results, but complete provenance. A trial manuscript now has to answer basic questions before a reviewer ever reaches the final page—what was planned, what changed, what was analyzed, and who could verify the underlying data. Without those answers, the manuscript does not clear the bar.
Editorial review will hinge on traceability, not presentation alone
The standards laid out here are especially demanding for interim analyses, complex biomarker studies, and AI-enabled clinical research, where analytic flexibility can blur the line between planned inference and exploratory pattern matching. By demanding protocol fidelity, prespecified reporting, and the appropriate guideline family for each study design, the journal reduces the room for interpretive drift. In a field where small analytic choices can alter the apparent weight of a result, those controls are not peripheral. They are the basis of the paper’s scientific value.
For authors, the message is unmistakable: if the trial was not registered before enrollment, if the protocol does not match the analysis, or if the data-sharing record is incomplete, the manuscript is exposed before review begins. For readers, the policy offers a clearer route to judging whether a published result deserves to influence care, guidelines, or further research. The most consequential limitation is also the most unforgiving one: a study that cannot meet these requirements will not be considered a full clinical trial report in this journal.