A reporting guideline helps authors describe what was done and found; it does not replace sound design or ethical approval. Select the current guideline and any relevant extension for the actual study design, using the EQUATOR Network library.

Guidelines by study design

• Randomized trials: CONSORT 2025 and relevant extensions; trial protocols should use SPIRIT 2025.
• Observational studies: STROBE and applicable extensions, including those for routinely collected or molecular epidemiological data.
• Systematic reviews and meta-analyses: PRISMA 2020 and the relevant extension; protocols should use PRISMA-P.
• Diagnostic accuracy: STARD and an applicable extension for the technology and study design.
• Clinical prediction models: TRIPOD+AI for regression or machine-learning model development and validation.
• Animal experiments: ARRIVE 2.0, including the Essential 10 and relevant Recommended Set items.
• Other eligible designs: use the appropriate EQUATOR-listed guideline; discuss an unusual article format with the editor.

Submission requirements

Upload a completed checklist with manuscript page or section references. Include the relevant flow diagram and explain items that do not apply. For trials and evidence syntheses, supply the protocol and explain changes from the prespecified methods. A checklist must reflect information actually present in the manuscript; it should not be completed by marking every item as satisfied without a location.

Cancer research and computational methods

Preclinical reports should state model provenance, authentication, sample-size reasoning, allocation, blinding, exclusions and the distinction between biological and technical replicates. Biomarker studies should define the intended use, sample handling, assay procedure and validation strategy. Computational oncology reports should identify data sources and accessions, preprocessing, software versions and the separation of development and validation data.

Use extensions relevant to AI, omics, imaging or other specialized methods. Complete reporting may require more than one checklist. Editors may request additional details when a standard checklist does not adequately describe the experiment or the clinical question. See Manuscript Preparation for the journal’s general requirements.