The CONSORT Statement (CONsolidated Standards Of Reporting Trials) is the reporting guideline for writing up the results of a randomized controlled trial (RCT). It defines the minimum information a published trial report must contain so readers, peer reviewers, and systematic reviewers can judge how the trial was conducted and whether its results are trustworthy.
CONSORT is maintained by an independent international group of trialists, statisticians, and journal editors, and is one of the reporting guidelines catalogued by the EQUATOR Network, alongside PRISMA for systematic reviews and STROBE for observational studies.
CONSORT 2025 has superseded CONSORT 2010
The version most researchers still refer to by habit — CONSORT 2010 (Schulz, Altman, and Moher, published in BMJ) — is no longer the current statement. On 14 April 2025, an updated CONSORT 2025 statement was published simultaneously in five major medical journals: The BMJ (2025;388:e081123), JAMA (2025;333:1998-2005), The Lancet, Nature Medicine (2025;31:1776-1783), and PLOS Medicine (2025;22(4):e1004587), led by Hopewell, Chan, Collins, and colleagues. A companion explanation and elaboration paper was published alongside it (BMJ 2025;388:e081124). Any trial report being prepared now should be checked against CONSORT 2025, not the 2010 checklist, unless a specific journal’s author instructions still explicitly reference the older version.
CONSORT 2025 expanded the checklist from 25 items to 30 items. The update added seven new items, revised three existing items, deleted one item (on generalizability), and folded in content that previously lived only in separate CONSORT extensions — including material on harms reporting, outcome reporting, and non-pharmacological treatments. The new items address areas the 2010 checklist did not require explicitly, including a data sharing statement, disclosure of authors’ financial and other conflicts of interest, patient and public involvement in the trial, eligibility criteria for the sites and individuals delivering an intervention, a dedicated harms-assessment item, an explicit definition of who is included in each analysis population, and further detail on how the intervention was actually delivered.
CONSORT 2010 vs. CONSORT 2025: what actually changed
Researchers who learned the checklist under CONSORT 2010 do not need to relearn it from scratch — most of the 2010 items carried forward, renumbered within CONSORT 2025’s six sections. The table below summarizes what changed, so a report prepared against the 2010 checklist can be brought up to the current standard.
| What changed | Detail |
|---|---|
| Total items | 25 items (CONSORT 2010) → 30 items (CONSORT 2025) |
| Structure | CONSORT 2010 did not group registration, data sharing, and funding/COI into one block. CONSORT 2025 consolidates them into a new, dedicated Open Science section. |
| Seven items added | Data sharing statement; financial and other conflicts of interest of manuscript authors (separate from funder role); patient and public involvement in the trial; eligibility criteria for the sites and individuals delivering the intervention; a dedicated harms-assessment item; an explicit definition of each analysis population; further detail on how the intervention was actually delivered in practice. |
| Three items revised | Existing 2010 items were reworded to sharpen exactly what must be reported, without changing their core topic. |
| One item deleted | The standalone item on generalizability/external validity was removed; related content is now expected within the discussion of limitations (item 30) rather than as its own checklist line. |
| Extension content folded in | Material that previously required consulting a separate CONSORT extension — harms reporting, outcome reporting detail, non-pharmacological treatment description — is now part of the core 30-item checklist. |
If a target journal’s author instructions still cite “CONSORT 2010” by name, check whether that is because the journal has not yet updated its instructions, or because it has a specific reason to still require the 2010 checklist; when in doubt, following CONSORT 2025 and noting the version used in the cover letter is the safer default, since CONSORT 2025 is the current statement endorsed by its authoring group.
What the 30-item checklist covers
CONSORT 2025 organizes its checklist into six sections. The overview below is a structural summary; the full item-by-item table further down reproduces every item, and researchers preparing a manuscript should still work from the actual checklist and the accompanying explanation and elaboration document on the official CONSORT website for exact wording.
- Title and abstract (2 items) — identification of the report as a randomized trial in the title, and a structured summary of trial design, methods, results, and conclusions.
- Open science (4 items) — a new section in the 2025 revision, consolidating trial registration, where the protocol and statistical analysis plan can be accessed, data sharing arrangements, and funding and conflicts of interest into one reporting block rather than scattering them through the report.
- Introduction (2 items) — scientific background and rationale, and the trial’s specific objectives or hypotheses.
- Methods — the largest section: patient/public involvement, trial design, eligibility criteria (now including criteria for sites and for the individuals delivering an intervention), interventions and how they were actually delivered, outcome definitions, harms assessment, sample size determination, randomization sequence generation, allocation concealment, blinding, and statistical methods, including how each analysis population is defined.
- Results (8 items) — participant flow through the trial, recruitment and trial dates, baseline demographic and clinical characteristics, numbers analyzed, outcome results with effect estimates and precision, ancillary and subgroup analyses, and harms.
- Discussion (2 items) — interpretation of the results in the context of existing evidence, and the trial’s limitations.
The full CONSORT 2025 checklist
This table reproduces the structure and topic of all 30 CONSORT 2025 checklist items, paraphrased into plain English, with a note on what a peer reviewer or editor is actually checking for against each one. Wording here is a paraphrase for clarity, not the verbatim checklist text — for the exact, citable wording used when completing a submission checklist, use the official checklist document from the CONSORT 2025 primary publication (Hopewell et al., BMJ 2025;388:e081123, and the companion explanation and elaboration paper, BMJ 2025;388:e081124).
| Section | Item | What must be reported | What a reviewer looks for |
|---|---|---|---|
| Title & Abstract | 1a | Identification as a randomised trial in the title | Checks the title itself states it is a randomised trial, not left to be inferred. |
| Title & Abstract | 1b | Structured summary of design, methods, results, and conclusions | Checks the abstract alone conveys accurate design and headline results. |
| Open Science | 2 | Trial registry name, registration number/URL, and date of registration | Checks registration predates enrollment and matches the reported outcomes. |
| Open Science | 3 | Where the full protocol and statistical analysis plan can be accessed | Checks for a working link/DOI, not a promise that details are “available on request.” |
| Open Science | 4 | Availability of de-identified participant data, statistical code, and materials | Checks the data-sharing statement specifies what is shared, with whom, and how — or explains why not. |
| Open Science | 5a | Sources of funding and the funder’s role in the trial | Checks whether the funder had any role in design, conduct, analysis, or the decision to publish. |
| Open Science | 5b | Financial and other conflicts of interest of the manuscript’s authors | Checks each author’s individual disclosure, not a single blanket statement. |
| Introduction | 6 | Scientific background and rationale | Checks the paper explains why the trial was needed, not only what it did. |
| Introduction | 7 | Specific objectives or hypotheses, including questions about harms | Checks that harm-related objectives are stated up front, not introduced later. |
| Methods | 8 | Patient and public involvement in designing, conducting, or reporting the trial | Checks whether patient/public partners contributed and how, or why they were not involved. |
| Methods | 9 | Trial design (e.g. parallel, factorial) and allocation ratio | Checks the design type and allocation ratio (e.g. 1:1) are stated explicitly. |
| Methods | 10 | Important changes to methods after the trial began | Checks for a clear account of protocol deviations or outcomes stopped/added mid-trial. |
| Methods | 11 | Settings and locations where data were collected | Checks the number and type of sites and the country/setting are stated. |
| Methods | 12a | Eligibility criteria for participants | Checks inclusion/exclusion criteria are specific enough to replicate the population. |
| Methods | 12b | Eligibility criteria for the sites and individuals delivering the intervention | New in 2025 — checks what qualified a site or clinician to deliver the intervention. |
| Methods | 13 | The interventions for each group, in replicable detail | Checks dose, timing, duration, and how the intervention was actually administered are specified. |
| Methods | 14 | Prespecified primary and secondary outcomes, and how/when measured | Checks reported outcomes match what was prespecified in the registry or protocol. |
| Methods | 15 | How harms were defined, identified, and assessed | Checks for a systematic harms-collection method, not an incidental mention. |
| Methods | 16a | Sample size determination, including every supporting assumption | Checks target difference, expected variability/event rate, power, and alpha are all stated. |
| Methods | 16b | Interim analyses and stopping guidelines | Checks whether a monitoring plan existed and, if the trial stopped early, why. |
| Methods | 17a | Method used to generate the random allocation sequence | Checks who generated the sequence and how (e.g. computer-generated). |
| Methods | 17b | Type of randomisation and any restriction (stratification, blocking) | Checks for detail beyond the bare word “randomised.” |
| Methods | 18 | Mechanism used to implement the sequence (allocation concealment) | Checks for the specific mechanism (e.g. sealed opaque envelopes, central randomisation) that hid the sequence until assignment. |
| Methods | 19 | Who generated, enrolled, and assigned participants | Checks these roles were kept separate to prevent foreknowledge of the next assignment. |
| Methods | 20a | Who was blinded after assignment (participants, care providers, assessors, analysts) | Checks blinding status is stated for every relevant role, not a vague “double-blind” label. |
| Methods | 20b | How blinding was achieved, and similarity of interventions if relevant | Checks whether blinding was verified or interventions were similar enough to sustain it. |
| Results | 21a | Statistical methods for comparing groups, including for harms | Checks the methods match the outcome types being compared. |
| Results | 21b | Definition of each analysis population (e.g. intention-to-treat, per-protocol) | Checks the analysis population is explicitly defined and consistently applied. |
| Results | 21c | How missing data were handled | Checks for a stated method rather than silence on dropout handling. |
| Results | 21d | Methods for additional analyses, prespecified vs. post hoc | Checks subgroup/sensitivity analyses are clearly labelled as planned or exploratory. |
| Results | 22a | Numbers randomised, receiving intended treatment, and analysed, per group | Checks these numbers are consistent with the flow diagram. |
| Results | 22b | Losses and exclusions after randomisation, with reasons, per group | Checks whether losses were balanced across arms or concentrated in one. |
| Results | 23a | Dates defining recruitment and follow-up | Checks the recruitment window and follow-up duration are both stated. |
| Results | 23b | Why the trial ended or was stopped, if applicable | Checks for an explicit explanation of any early stop. |
| Results | 24a | What was actually delivered and how closely it matched the plan (fidelity) | Checks for reported adherence/fidelity data, not just the planned protocol restated. |
| Results | 24b | Concomitant care received by each group during the trial | Checks for co-interventions that could confound the result if unevenly distributed. |
| Results | 25 | Baseline demographic and clinical characteristics table | Checks the table exists and looks for imbalances that could explain outcome differences. |
| Results | 26 | Results, effect estimates, and precision for each outcome | Checks effect sizes and confidence intervals are reported, not p-values alone. |
| Results | 27 | All important harms or unintended effects in each group | Checks harms reporting is as complete and structured as efficacy reporting. |
| Results | 28 | Results of any other analyses performed | Checks these are reported even when null, and flagged as exploratory. |
| Discussion | 29 | Interpretation balancing benefits, harms, and other evidence | Checks the conclusion does not overreach beyond what the data support. |
| Discussion | 30 | Trial limitations: bias, imprecision, and (if relevant) multiplicity | Checks limitations are specific to this trial, not generic boilerplate. |
Items with a letter suffix (e.g. 12a/12b, 17a/17b) are sub-parts of a single numbered checklist item that must both be addressed, not two separate items.
Several of the Methods and Results items — most notably randomization sequence generation (17a/17b) and allocation concealment (18) — are commonly conflated in practice but are reported as separate checklist items, because they address different sources of bias: sequence generation concerns whether the assignment order was genuinely unpredictable, while allocation concealment concerns whether that sequence was hidden from whoever enrolled participants until the moment of assignment.
The CONSORT flow diagram: four stages and a copyable template
Alongside the checklist, CONSORT requires a flow diagram documenting how participants moved through the trial. The diagram exists because attrition and exclusions happening unevenly across arms are a recognized source of bias that a results narrative alone can obscure; a reader can trace the diagram to see exactly how the analyzed sample was arrived at, arm by arm. The diagram has four stages:
| Stage | What it records |
|---|---|
| 1. Enrollment | Number assessed for eligibility; number excluded, broken down by reason (did not meet inclusion criteria, declined to participate, other reasons); number randomized. |
| 2. Allocation | For each arm: number allocated to the intervention; number who actually received the allocated intervention; number who did not receive it, with reasons. |
| 3. Follow-Up | For each arm: number lost to follow-up, with reasons; number who discontinued the intervention, with reasons. |
| 4. Analysis | For each arm: number included in the final analysis; number excluded from the analysis, with reasons. |
A minimal, fill-in-the-blank version of the diagram, in a format that can be copied into a manuscript draft or diagramming tool:
ENROLLMENT
Assessed for eligibility (n = ___)
Excluded (n = ___)
- Not meeting inclusion criteria (n = ___)
- Declined to participate (n = ___)
- Other reasons (n = ___)
Randomized (n = ___)
ALLOCATION
Arm A: Allocated (n = ___) -> Received intervention (n = ___) / Did not receive (n = ___, reasons: ___)
Arm B: Allocated (n = ___) -> Received intervention (n = ___) / Did not receive (n = ___, reasons: ___)
FOLLOW-UP
Arm A: Lost to follow-up (n = ___, reasons: ___); Discontinued (n = ___, reasons: ___)
Arm B: Lost to follow-up (n = ___, reasons: ___); Discontinued (n = ___, reasons: ___)
ANALYSIS
Arm A: Analyzed (n = ___); Excluded from analysis (n = ___, reasons: ___)
Arm B: Analyzed (n = ___); Excluded from analysis (n = ___, reasons: ___)
Most journals expect the completed diagram rendered as a numbered figure (typically Figure 1) within the manuscript itself, not only as a supplementary file — see the submission workflow below.
CONSORT extensions
Beyond the core statement, CONSORT maintains a family of extensions adapting the checklist to trial designs and reporting situations the core checklist does not fully address on its own — including cluster-randomized trials, non-inferiority and equivalence trials, pragmatic trials, adaptive designs, crossover trials, and factorial designs, as well as extensions focused on specific reporting elements such as harms and abstracts. CONSORT 2025 folded some previously extension-only content (harms, outcomes, non-pharmacological treatment detail) into the core checklist itself, but design-specific extensions remain relevant for trials with those specific designs and should be consulted in addition to, not instead of, the core 2025 checklist.
Why journals require it
CONSORT’s practical authority comes from journal adoption rather than any regulatory mandate. It is endorsed by a large number of biomedical journals worldwide, and the International Committee of Medical Journal Editors (ICMJE) recommends that journals require authors to follow the relevant EQUATOR Network reporting guideline for their study type — CONSORT for randomized trials, PRISMA for systematic reviews, STROBE for observational studies. In practice this usually means a submitting author is asked to supply a completed CONSORT checklist, keyed to manuscript page and line numbers, plus a flow diagram, at submission. Research comparing reporting quality before and after journal endorsement consistently finds an association with more complete reporting, though a causal effect is difficult to establish cleanly since journals that endorse CONSORT may also differ from non-endorsing journals in other ways.
Submitting your CONSORT checklist to a journal
The checklist and flow diagram are prepared alongside the manuscript, then submitted as part of the same package. The general workflow, common across most biomedical journals that require CONSORT:
- Download the current fillable checklist from the official CONSORT website, matching the version the target journal expects — CONSORT 2025 unless a specific journal’s instructions still name CONSORT 2010.
- Complete it against the manuscript, noting the page and line number where each item is addressed; if an item genuinely does not apply to the trial’s design, state why rather than leaving it blank.
- Upload it as a distinct supplementary file during online submission — most manuscript submission systems list “CONSORT Checklist” as its own file-type option, separate from the main manuscript and from other supplementary material.
- Include the flow diagram as a numbered figure in the manuscript itself (commonly Figure 1), not only inside the supplementary checklist file.
- Check whether the journal also asks for a compliance statement in the cover letter or a submission questionnaire confirming CONSORT was followed; requirements vary by journal, so the specific author instructions for the target journal are the authoritative source for that journal’s exact process.
How this fits into a trial’s workflow
For a research administrator or trial coordinator, CONSORT is best treated as a manuscript-preparation and quality-assurance tool that should be anticipated well before the results-writing stage, not addressed retroactively:
- Trial registration (a separate requirement from CONSORT itself — see clinical trial registration and results-reporting compliance) and the statistical analysis plan should be finalized and accessible before results reporting, since CONSORT 2025’s Open Science section expects the report to state where both can be found.
- Participant-flow data (screened, randomized, allocated, followed up, analyzed, and the reasons for any loss at each stage) is far easier to compile accurately if it is tracked as the trial runs, rather than reconstructed afterward from site records.
- Harms and adverse-event data, now an explicit checklist item rather than an optional add-on, needs the same rigor in collection and adjudication as the trial’s primary efficacy outcomes if it is to be reported to CONSORT’s standard.
- If the trial uses a design covered by a CONSORT extension (cluster randomization, non-inferiority, a pragmatic design, and so on), identify the applicable extension early so protocol-level decisions capture what that extension will require to be reported later.
CONSORT, PRISMA, and STROBE: which guideline applies
These three EQUATOR-catalogued guidelines are often mentioned together because they cover adjacent stages of the same evidence base, but each applies to a different kind of report and they are not interchangeable:
- CONSORT applies to reports of individual randomized controlled trials.
- PRISMA applies to systematic reviews and meta-analyses that synthesize evidence across multiple studies, which may themselves include CONSORT-reported trials. See CASRAI’s PRISMA and systematic review methodology guide for the full treatment.
- STROBE applies to observational studies (cohort, case-control, and cross-sectional designs) that do not involve random allocation to an intervention.
A trial team should not need to choose between these — the guideline is determined by the study design being reported, not by preference. For the broader category these all belong to, see CASRAI’s reporting guidelines overview, and for CASRAI’s wider clinical-research coverage, see the clinical research pillar.
Frequently asked questions
What does CONSORT stand for?
CONsolidated Standards Of Reporting Trials.
Is CONSORT 2010 still valid, or has it been replaced?
CONSORT 2025, published 14 April 2025 in The BMJ, JAMA, The Lancet, Nature Medicine, and PLOS Medicine, is the current statement and expands the checklist from 25 to 30 items. CONSORT 2010 is superseded; check the specific target journal’s current author instructions if there is any doubt about which version it expects. See the 2010-vs-2025 comparison table above for what specifically changed.
Does CONSORT apply to every clinical trial?
The core statement applies specifically to two-group, parallel-design randomized controlled trials. Trials with other designs — cluster-randomized, non-inferiority or equivalence, pragmatic, adaptive, crossover, or factorial — should also consult the relevant CONSORT extension for that design.
Do all journals require a completed CONSORT checklist at submission?
Not universally, but it is a common requirement among biomedical journals that publish trial results, particularly ICMJE-affiliated titles, and is typically requested as a checklist keyed to manuscript page/line numbers alongside the flow diagram.
What is the difference between the CONSORT checklist and the CONSORT flow diagram?
The checklist is a list of items the manuscript text itself must address (design, methods, results, and so on); the flow diagram is a separate visual accounting of how many participants entered, moved through, and were ultimately analyzed in each arm of the trial.
What are the four stages of the CONSORT flow diagram?
Enrollment, Allocation, Follow-Up, and Analysis. See the flow diagram section above for what each stage records and a copyable fill-in template.
Where do I submit a completed CONSORT checklist?
As a supplementary file during the journal’s online manuscript submission, typically alongside a flow diagram included as a numbered figure in the manuscript itself. See “Submitting your CONSORT checklist to a journal” above for the full workflow.
