nejm-fit

Clinical Significance Fit (nejm-fit)

Why this is skill #1

NEJM triages the large majority of submissions to rejection without external review. The gate is not "is the study sound" — it is "would this change clinical practice, and is the evidence definitive enough to justify that change." A well-conducted but narrow study is desk-rejected. Run this before investing in prose.

When to trigger

• Before drafting, to decide if NEJM is even the right venue.
• When a co-author says "this is an NEJM paper" and you need a sober second opinion.
• When choosing between NEJM, Lancet, JAMA, a specialty journal (e.g., Circulation, Blood, JCO), and NEJM Evidence.

The three gates (all must hold)

NEJM weighs three things together. A paper passes only if it clears all three:

1. Clinical importance — does it address a question clinicians and patients actually face, with an outcome that matters (mortality, major morbidity, function, quality of life — not just a surrogate)?
2. Methodological rigor — is the design strong enough that the result is believable and not likely to be overturned (adequately powered RCT, rigorous observational design with confounding addressed)?
3. Generalizability — do the findings extend beyond a single center / narrow population to the broad practice community NEJM serves?

A large RCT with a surrogate endpoint can still fail gate 1. A striking finding from one underpowered single-center study fails gates 2–3.

Significance ladder (weak → strong)

1. Case report / small case series. (Weak — correspondence or specialty journal.)
2. Mechanistic or early-phase finding without clinical outcomes. (Specialty / translational journal.)
3. Single rigorous study extending known effects to a new population. (Borderline — JAMA/Lancet/specialty.)
4. Definitive RCT or landmark study answering a practice question with a hard outcome. (Strong.)
5. Practice-changing trial that resolves a controversy or sets a new standard of care. (Strongest.)

If you cannot place the work at rung 4+, NEJM is a long shot — be honest with the user and name the realistic target.

Fatal desk-reject triggers

• Outcome is a surrogate (lab value, imaging marker) with no patient-important endpoint.
• Underpowered for the primary outcome, or the primary outcome was changed post hoc.
• Single-center, narrow population, with no claim to generalizability.
• Not prospectively registered for a trial (an ICMJE deal-breaker — see nejm-study-design).
• Over-claiming: causal language on observational data, or a subgroup result sold as the main finding.
• Incremental over the authors' own prior trial with no new practice implication.

Venue routing

Situation	Recommend
Definitive, practice-changing, generalizable RCT/landmark study	NEJM (Original Article)
Rigorous and important, but global-health or broad public-health framing	The Lancet
Strong clinical trial/study, large general-medicine audience	JAMA
Methodologically strong, fits open-science/registration ethos	BMJ
Important to one specialty, not broad practice	specialty journal (Circulation, Blood, JCO, …)
Solid but not top-tier general impact; pragmatic/methods focus	NEJM Evidence / specialty
Early-phase / mechanism / surrogate only	translational or specialty journal

Output format

【Three gates】 clinical importance / rigor / generalizability — pass or fail each, one line
【Significance rung】 1–5 + one-line justification
【Outcome type】 patient-important / surrogate-only → flag if surrogate
【Fatal triggers present】 [...]
【Recommended venue】 NEJM / Lancet / JAMA / BMJ / specialty / NEJM Evidence
【If staying with NEJM, the single sentence of practice-changing impact】 "..."
【Next】 nejm-study-design (if pass) | reconsider venue (if fail)

Anti-patterns

• Do not rationalize a narrow result into "practice-changing" with adjectives — editors discount adjectives.
• Do not confuse a statistically significant surrogate endpoint with clinical importance.
• Do not let sample size alone stand in for rigor — a large biased study is still biased.
• Do not let sunk cost ("the trial took five years") drive the venue decision.