From lancet-skills
Generates and validates Lancet-compliant clinical display items: Table 1 baseline characteristics, Kaplan-Meier curves, forest plots, CONSORT/PRISMA flow diagrams, and multi-country study maps.
How this skill is triggered — by the user, by Claude, or both
Slash command
/lancet-skills:lancet-figures-tablesThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- The manuscript lacks a Table 1, a flow diagram, or a survival/forest plot the design calls for.
lancet-ethics.lancet-statistics.lancet-reporting. It is a required display item; numbers must reconcile with Table 1 and the analysis populations.The Lancet's clinical readership reads tables and figures before the prose, so a display item that hides uncertainty or contradicts the text is an immediate credibility problem. The statistical reviewer checks that every effect carries a 95% CI, that Kaplan–Meier curves expose the thinning denominator via a numbers-at-risk row, and that subgroup forests report an interaction test, not within-subgroup P values. Desk-reject-adjacent problems: effects with no CI; baseline P values in an RCT Table 1; survival curves with no numbers at risk; a subgroup figure with no interaction test; a multi-country study with no map. Confirm the display-item allowance and figure specifications in the journal's author guidelines.
A hypothetical cardiovascular RCT, composite MACE outcome over 36 months.
Kaplan-Meier (illustrative): 36-mo event rate 12.4% vs 16.8%
Hazard ratio 0.73 (95% CI 0.62-0.86), log-rank p=0.0002
Numbers at risk (0/36 mo): 2 050/1 690 vs 2 048/1 588
Absolute risk difference 4.4 pp; NNT ~23 over 3 years (illustrative).
Subgroup forest: 6 pre-specified, overall HR 0.73 (0.62-0.86), interaction p=0.41.
Table 1: arm columns, mean (SD) / n (%), NO baseline P column.
The numbers-at-risk row shows the curves stay reliable, the NNT gives the magnitude the relative HR hides, and the interaction P keeps the subgroup figure honest.
【Table 1】 present? arms × characteristics, correct summaries, no baseline P in RCT, equity vars? → gaps
【Kaplan–Meier】 needed? numbers-at-risk + HR(95% CI) present? yes/no/NA
【Forest plot】 subgroups (interaction p) / meta-analysis (pooled + heterogeneity)? yes/no/NA
【Flow diagram】 CONSORT/PRISMA present + reconciles? yes/no
【Map】 multi-country study has a map? yes/no/NA
【CIs shown on all estimates?】 yes/no
【Next】 lancet-research-in-context
npx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin lancet-skillsBuilds NEJM clinical display items correctly — Table 1 with standardized differences, Kaplan-Meier curves with numbers-at-risk, forest plots for subgroups, and CONSORT flow diagram.
Designs and audits figures and tables for JAMA manuscripts, including flow diagrams, baseline tables, and main-result exhibits with 95% CIs.
Builds PRISMA flow diagrams, forest plots, funnel/bias plots, moderator bubble plots, and APA 7th-edition summary tables for Psychological Bulletin meta-analysis manuscripts.