Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› lancet-reporting
lancet-reporting
GitHub用于根据研究类型选择正确的EQUATOR报告指南（如CONSORT、PRISMA等），生成并附带完整的检查清单及必需的研究流程图，确保符合柳叶刀期刊的提交要求。
Trigger Scenarios
确定研究类型后需匹配报告清单
稿件缺少参与者流程图
审稿人要求提供带页码行号的完整报告清单
稿件数据与流程图不一致需核对
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill lancet-reporting -g -y
SKILL.md
Frontmatter
{
"name": "lancet-reporting",
"description": "Use to select and apply the correct EQUATOR reporting guideline for a Lancet manuscript — CONSORT for RCTs (with the mandatory flow diagram), STROBE for observational studies, PRISMA for systematic reviews\/meta-analyses, SPIRIT for protocols — and to attach the completed checklist and study-flow diagram."
}
Reporting Guidelines & Flow Diagrams (lancet-reporting)
When to trigger
- The study type is settled and you need the matching reporting checklist.
- There is no participant flow diagram (CONSORT/PRISMA) in the manuscript.
- A reviewer or editor asks for a completed reporting checklist with page/line references.
- The manuscript mixes endpoints/numbers that the flow diagram should reconcile.
Pick the guideline by study type (via EQUATOR)
The Lancet requires the appropriate EQUATOR Network reporting guideline and a completed checklist on submission.
| Study type | Guideline | Mandatory diagram |
|---|---|---|
| Randomized controlled trial | CONSORT (+ extensions: cluster, non-inferiority, pragmatic, harms, PRO) | CONSORT flow diagram (enrollment → allocation → follow-up → analysis) |
| Observational (cohort, case-control, cross-sectional) | STROBE | Participant flow / selection diagram |
| Systematic review / meta-analysis | PRISMA (+ PRISMA-P for protocols) | PRISMA flow diagram (records → screened → included) |
| Trial / review protocol | SPIRIT (trials) / PRISMA-P (reviews) | — |
| Diagnostic accuracy | STARD | Patient flow |
| Prediction model | TRIPOD | — |
| Quality improvement | SQUIRE | — |
| Economic evaluation | CHEERS | — |
The flow diagram is mandatory for RCTs and systematic reviews
- CONSORT flow diagram present for every RCT: numbers assessed for eligibility, excluded (with reasons), randomized, allocated, lost to follow-up/discontinued (with reasons), and analyzed in each arm.
- PRISMA flow diagram present for every systematic review/meta-analysis: records identified, duplicates removed, screened, full-text assessed, excluded (with reasons), and included.
- Every number in the flow diagram reconciles with Table 1, the analysis populations, and the abstract.
Checklist discipline
- Complete the official checklist for the chosen guideline.
- Map each item to a page and line in the manuscript (editors and reviewers check this).
- Use the correct extension (e.g., CONSORT cluster for cluster-randomized trials; CONSORT harms for adverse-event reporting; CONSORT-PRO for patient-reported outcomes).
- Submit the checklist as a supplementary file.
Equity-relevant reporting
- Apply equity-focused extensions where relevant (e.g., PRISMA-Equity, CONSORT-Equity considerations) and the PROGRESS-Plus framework for reporting equity-relevant participant characteristics.
- Coordinate sex/gender and race/ethnicity reporting with
lancet-ethics(SAGER guidelines).
What Lancet editors and reviewers expect from reporting compliance
A completed checklist is a screening gate at The Lancet, not a formality. Editors and the statistical reviewer cross-read the flow diagram against Table 1 and the analysis populations; a discrepancy reads as a transcription error or selective reporting, and unexplained attrition is where bias hides. Desk-reject-adjacent patterns: no flow diagram for an RCT or review (effectively mandatory); base CONSORT where a cluster/non-inferiority extension fits; a checklist that says "throughout" instead of page/line; and numbers that do not reconcile. Confirm any current formatting requirement against the journal's author guidelines.
Worked micro-example (illustrative numbers — not real data)
A hypothetical multi-country, individually randomised trial of an oral therapy versus placebo across 14 sites in 5 low- and middle-income countries.
Guideline: CONSORT 2010 + CONSORT-Harms; checklist page/line-mapped as a supplement.
CONSORT flow diagram (illustrative):
Assessed 5 312 -> Excluded 1 108 (ineligible 842; declined 201; other 65)
Randomised 4 204 -> intervention 2 102 / placebo 2 102
Lost to follow-up: intervention 51, placebo 58 (reasons listed)
Analysed (ITT): 2 102 / 2 102
Reconciliation: 4 204 = Table 1 total = abstract Findings denominator. PASS.
The reporting pass is the reconciliation: 4 204 appears identically in the diagram, Table 1, and the abstract, and every exclusion and dropout carries a reason.
Reviewer-pushback patterns and the venue-specific fix
- "The reporting checklist is incomplete / not page-referenced." → Complete the official EQUATOR checklist and map each item to a page and line; resubmit as a supplement.
- "This is a cluster-randomised trial but reported under standard CONSORT." → Switch to CONSORT-Cluster with cluster-level flow and intracluster correlation reporting.
- "Flow-diagram numbers do not match the analysis populations." → Rebuild the diagram so randomised, analysed, and Table 1 totals reconcile; explain any gap.
- "Equity-relevant characteristics are not reported." → Add PROGRESS-Plus variables and a PRISMA-Equity/CONSORT-Equity lens; coordinate with
lancet-ethics.
Output format
【Study type】 RCT / observational / systematic review / protocol / other
【Guideline + extension】 CONSORT(+ext) / STROBE / PRISMA(+ext) / SPIRIT / STARD / TRIPOD / ...
【Flow diagram】 present & required? CONSORT / PRISMA / N/A → gaps
【Flow numbers reconcile】 with Table 1 + analysis populations + abstract? yes/no
【Checklist】 completed + page/line mapped? yes/no
【Equity reporting】 PROGRESS-Plus / equity extension applied where relevant? yes/no
【Next】 lancet-statistics
Anti-patterns
- Do not submit an RCT without a CONSORT flow diagram, or a systematic review without a PRISMA one.
- Do not leave flow-diagram numbers that disagree with the analysis populations or Table 1.
- Do not submit a checklist with "see manuscript" instead of specific page/line references.
- Do not use the base guideline when an extension fits the design better (cluster, non-inferiority, pragmatic, harms).
Version History
- 1839142 Current 2026-07-05 14:01
