jama-figures-tables

Figures & Tables (jama-figures-tables)

When to trigger

• The RCT/review lacks its required flow diagram
• Tables are cluttered, over-numerous, or exceed the journal's exhibit allowance
• The baseline-characteristics table is missing or misused
• A figure shows p-values but not effect sizes / confidence intervals

The exhibits a JAMA paper usually needs

| Exhibit | Purpose |

|-------------------------------------------|--------------------------------------------------------|

| Flow diagram (CONSORT/PRISMA/STARD) | Participant or study-selection accounting |

| Table 1 — baseline characteristics | Describe groups; show comparability (not significance) |

| Main-outcome table | Primary/secondary effects with 95% CIs |

| Forest plot (meta-analysis / subgroups) | Effect sizes + CIs across studies/subgroups |

| Kaplan-Meier / time-to-event figure | Survival or event curves with numbers at risk |

Keep the count within JAMA's current figure/table limit — verify the exact number on the Instructions for Authors page rather than assuming.

Conventions JAMA expects

• Flow diagram is mandatory for RCTs (CONSORT) and systematic reviews (PRISMA). Account for every participant/record: screened, excluded (with reasons), enrolled, analyzed.
• Table 1 shows comparability, not p-values. For RCTs, comparing baseline groups with significance tests is discouraged; report characteristics and let randomization speak.
• Every estimate carries its 95% CI in figures and tables, not bare p-values.
• Self-contained exhibits. Each table/figure stands alone: full title, defined abbreviations in the footnote, units, denominators (n/N), and the analysis population.
• Kaplan-Meier curves include numbers at risk beneath the time axis.
• Forest plots order studies sensibly, show weights and the pooled estimate with its CI, and report heterogeneity.
• No chartjunk. Avoid 3-D bars, truncated axes that exaggerate effects, and dual axes that mislead.
• Color and accessibility. Use colorblind-safe palettes; ensure the figure reads in grayscale.

Checklist

• [ ] Required flow diagram present and reconciles all participants/records
• [ ] Table 1 reports characteristics without inappropriate significance testing (RCT)
• [ ] Every effect estimate shows its 95% CI
• [ ] Each exhibit is self-contained (title, footnotes, abbreviations, units, n/N)
• [ ] KM curves show numbers at risk; forest plots show weights + pooled CI + heterogeneity
• [ ] Total figures + tables within the journal's limit (verify current number)
• [ ] Axes honest (no truncation that exaggerates); palette colorblind-safe
• [ ] No duplication of the same data across a table and a figure

Anti-patterns

• Submitting an RCT or review with no flow diagram
• P-value-laden Table 1 implying randomization "failed"
• Bars or curves with no confidence intervals
• Truncated y-axis that visually inflates a small effect
• KM plot without numbers at risk
• The same numbers shown in both a table and a figure
• Abbreviations undefined in the footnote

Worked example: auditing the main exhibits (illustrative)

Vignette (illustrative): a multicenter randomized clinical trial, N = 2,200 adults, time-to-event primary outcome. For the Journal of the American Medical Association, the exhibit set is a CONSORT flow diagram reconciling all 2,200 participants (screened, randomized, analyzed, lost), a Table 1 of baseline characteristics shown without between-group significance tests, a main-outcome table giving the hazard ratio with its 95% CI plus the absolute risk difference, and a Kaplan-Meier figure with numbers at risk beneath the time axis. A bar chart of the primary outcome with no confidence interval, or a truncated y-axis that visually inflates a small effect, is the kind of exhibit a JAMA reviewer sends back.

Reviewer pushback and the JAMA fix

• "RCT submitted with no flow diagram." Fix: add the CONSORT diagram and reconcile every participant with reasons for exclusion.
• "Table 1 reports p-values (RCT)." Fix: drop the significance tests; show characteristics and let randomization speak.
• "Figure shows p-values, not effect sizes." Fix: display the estimate with its 95% CI on every exhibit.

Calibration anchors (hedge where uncertain): the mandatory flow diagram, p-value-free Table 1 for RCTs, estimate-plus-CI on every exhibit, and numbers-at-risk on KM curves are durable; the exact figure/table allowance is volatile — confirm against current author guidelines.

Exhibit pass for JAMA

Run this as a concrete capability pass. First lock the clinical question, patient population, estimand or endpoint, safety/ethics issue, and reporting checklist; then test whether the manuscript addresses clinical reviewers who ask whether the evidence changes patient care, policy, or medical decision-making while satisfying reporting standards.

• Primary move: For every figure or table, state the patient/study object, denominator, uncertainty display, and the clinical claim it supports.
• Decision ledger: return claim / evidence / blocker / next edit rows so the next pass can patch the manuscript directly.
• Sibling comparison: compare against NEJM for field-changing clinical medicine, Lancet for global-health breadth, specialty journals for narrower clinical 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

【Flow diagram】present + reconciles / missing
【Table 1 appropriate】yes / no
【All estimates show 95% CI】yes / no
【Exhibit count vs limit】X (limit per author instructions)
【Self-contained exhibits】yes / fixes: ...
【Issues to fix】...
【Next skill】jama-structured-abstract