Reporting Guidelines (nejm-reporting)

When to trigger

• You must pick the reporting checklist that matches the study type.
• A trial write-up has no CONSORT participant flow diagram.
• A reviewer or editor will ask for a completed reporting checklist with page/line references.
• An observational study or systematic review needs STROBE / PRISMA compliance.

Match the study type to the guideline (EQUATOR Network)

The EQUATOR Network indexes reporting guidelines by study type. The core ones NEJM cares about:

Match the extension, not just the base guideline: a cluster-randomized or non-inferiority trial uses the corresponding CONSORT extension.

The CONSORT participant flow diagram (mandatory for RCTs)

For an RCT, the flow diagram is not optional. It traces participants through four stages:

1. Enrollment — assessed for eligibility; excluded (with reasons); randomized.
2. Allocation — allocated to each arm; received allocated intervention or not.
3. Follow-up — lost to follow-up and discontinued (with reasons), per arm.
4. Analysis — analyzed; excluded from analysis (with reasons), per arm.

Numbers must reconcile with Table 1, the analysis populations, and the text. Mismatched denominators across the flow diagram, Table 1, and results are a frequent reviewer catch.

What to deliver for each study type

• RCT → completed CONSORT checklist (item → page/line), the flow diagram, and confirmation that registration number, protocol, and SAP are present (see nejm-study-design).
• Observational → completed STROBE checklist; define cohort entry, follow-up, and how confounders were handled.
• SR/MA → completed PRISMA checklist, the selection flow diagram, the search strategy, and (ideally) a PROSPERO registration number.

Running the checklist so it survives revision

Treat the completed checklist as a living index, not a one-time export:

• Fill each item with a page + line reference plus a short quote that re-locates the item after re-pagination.
• Mark "not applicable" only with a one-line justification; silent N/A rows read as unread items.
• Re-run the page/line pass after every revision round; one author owns the checklist.

Worked micro-example — one checklist row (before → after)

• Before (too thin): Item 8a — Sequence generation: "Methods"
• After: Item 8a — Sequence generation: p. 6, lines 112–115 — "randomization was performed centrally in permuted blocks, stratified by site"

Reconciliation arithmetic (do it on paper)

Force the sums across the four stages: assessed − excluded = randomized; randomized = the sum of the arms; each arm's allocated n − (lost + discontinued) traces to the analyzed n with every exclusion named, then check the same numbers against Table 1 headers and primary-outcome denominators. Fictional example: 1042 assessed − 562 excluded = 480 randomized = 241 + 239; if Table 1 shows 240 + 239, the missing participant must be explained in the diagram.

Operating pass for New England Journal of Medicine

Use this as a second-pass capability check. First lock the clinical question, population, endpoint, effect size, safety signal, and practice implication; then test whether the manuscript addresses clinical-medicine reviewers who expect practice-changing evidence, patient relevance, safety, and exact reporting discipline.

• Primary move: Return a claim-evidence-risk ledger; every recommendation must point to a manuscript location or missing artifact.
• Decision ledger: return claim / evidence / blocker / next edit rows so the next pass can patch the manuscript directly.
• Neighbor test: compare against JAMA for broad clinical medicine, Lancet for global-health/public-health reach, specialty journals for narrower disease domains; if the neighboring outlet has the stronger audience claim, recommend re-routing before polishing.
• Verification floor: before submission-ready advice, re-open resources/official-source-map.md for volatile rules and name the one unresolved fact that could change the recommendation.

Output format

【Study type】 ...
【Guideline + extension】 CONSORT (+ cluster/non-inferiority?) / STROBE / PRISMA / SPIRIT / CARE / STARD
【Required diagram】 CONSORT flow / PRISMA selection / STROBE eligibility — present? yes/no
【Checklist status】 completed with page/line refs? yes/no — gaps: [...]
【Number reconciliation】 flow diagram ↔ Table 1 ↔ analysis populations consistent? yes/no
【Next】 nejm-writing

Anti-patterns

• Do not submit an RCT without a CONSORT participant flow diagram.
• Do not use the base CONSORT checklist for a cluster or non-inferiority trial — use the extension.
• Do not let the flow-diagram denominators disagree with Table 1 or the analysis populations.
• Do not treat the checklist as paperwork — reviewers verify items against the actual text.