Build publication-quality figures and tables
from TriNetX in seconds.

Launch OutcomePlot See how it works →
OutcomePlot  ·  Roux-en-Y gastric bypass vs Sleeve Gastrectomy  ·  PSM on 90 covariates  ·  n = 19,529 each
Characteristic Before PSM After PSM
Roux-en-Y gastric bypassSleeve GastrectomySMD Roux-en-Y gastric bypassSleeve GastrectomySMD
Total n 19,66556,015 19,52919,529
Demographics
Age at Index (years, mean ± SD)45.2 ± 12.243.8 ± 11.90.12045.2 ± 12.245.4 ± 12.10.015
Female, n (%)15,934 (81.0%)45,503 (81.2%)0.00515,824 (81.0%)15,692 (80.4%)0.017
Male, n (%)3,731 (19.0%)10,512 (18.8%)0.0053,705 (19.0%)3,837 (19.6%)0.017
White, n (%)14,163 (72.0%)35,304 (63.0%)0.19314,052 (72.0%)14,119 (72.3%)0.008
Black or African American, n (%)3,302 (16.8%)14,026 (25.0%)0.2043,298 (16.9%)3,207 (16.4%)0.013
Asian, n (%)72 (0.4%)252 (0.4%)0.01372 (0.4%)66 (0.3%)0.005
Hispanic or Latino, n (%)3,236 (16.5%)6,773 (12.1%)0.1253,202 (16.4%)3,295 (16.9%)0.013
Clinical
Body mass index (kg/m², mean ± SD)43.7 ± 8.043.5 ± 7.80.02943.7 ± 8.043.6 ± 7.90.020
Systolic blood pressure (mmHg, mean ± SD)129.2 ± 16.6129.3 ± 16.70.010129.1 ± 16.6129.1 ± 16.40.004
Diastolic blood pressure (mmHg, mean ± SD)78.0 ± 11.177.9 ± 11.40.00278.0 ± 11.177.8 ± 11.10.018
Laboratory
HbA1c (%, mean ± SD)6.3 ± 1.56.2 ± 1.40.0786.3 ± 1.56.2 ± 1.40.052
Creatinine (mg/dL, mean ± SD)0.8 ± 1.00.8 ± 0.60.0110.8 ± 1.00.8 ± 0.50.005
Total cholesterol (mg/dL, mean ± SD)181.5 ± 38.0181.7 ± 38.10.004181.5 ± 37.9181.5 ± 39.00.002
LDL cholesterol (mg/dL, mean ± SD)107.1 ± 32.5107.4 ± 32.20.008107.2 ± 32.5106.8 ± 32.40.012
HDL cholesterol (mg/dL, mean ± SD)47.0 ± 13.546.6 ± 14.30.02347.0 ± 13.546.3 ± 13.70.050
Triglycerides (mg/dL, mean ± SD)144.0 ± 104.8141.3 ± 107.90.026143.8 ± 105.3147.1 ± 119.00.029
Outcome
Events / N(Roux-en-Y gastric bypass)
Events / N(Sleeve Gastrectomy)
RD per 1000 PY(95% CI)
NNT / NNH (PY)(95% CI)
Hazard Ratio
HR (95% CI)
E-value(point / CI)
MALO composite
1,047 / 19,529
708 / 19,529
3.63 (2.85, 4.41)
NNH: 276 (227-351)
1.58 (1.44–1.74)
2.55 / 2.24
Hyperlipidemia
6,650 / 19,529
7,876 / 19,529
-7.99 (-10.21, -5.77)
NNT: 125 (98-173)
0.82 (0.79–0.84)
1.75 / 1.65
Ascites
622 / 19,529
317 / 19,529
3.11 (2.54, 3.68)
NNH: 322 (272-394)
2.08 (1.82–2.38)
3.58 / 3.04
Cirrhosis
218 / 19,529
178 / 19,529
0.47 (0.11, 0.84)
NNH: 2107 (1185-9460)
1.29 (1.06–1.58)
1.91 / 1.32
Liver cancer
16 / 19,529
15 / 19,529
0.02 (-0.09, 0.12)
NNH: 58630
1.11 (0.55–2.26)
1.47 / 1.00
Encephalopathy
339 / 19,529
265 / 19,529
0.85 (0.39, 1.30)
NNH: 1181 (768-2560)
1.40 (1.19–1.64)
2.14 / 1.67
Varices
34 / 19,529
26 / 19,529
0.09 (-0.05, 0.23)
NNH: 11086
1.44 (0.86–2.41)
2.24 / 1.00
Portal hypertension
86 / 19,529
53 / 19,529
0.35 (0.13, 0.56)
NNH: 2895 (1771-7933)
1.75 (1.24–2.47)
2.90 / 1.79
All-cause mortality
576 / 19,529
370 / 19,529
2.18 (1.60, 2.75)
NNH: 460 (364-624)
1.70 (1.49–1.93)
2.78 / 2.34
0.5 0.75 1 1.5 2 3 4
← Favors Roux-en-Y gastric bypassFavors Sleeve Gastrectomy →
OutcomePlot  ·  2026-04-22  ·  TriNetX Compare Outcomes  ·  Roux-en-Y gastric bypass (n=19,665→19,529); Sleeve Gastrectomy (n=56,015→19,529)  ·  PSM on 90 covariates  ·  median follow-up 1899/2058 d (Roux-en-Y gastric bypass/Sleeve Gastrectomy)
MALO composite
Cumulative incidence, % 0 5 10 15 0 1 2 3 4 5 Time, years Roux-en-Y gastric bypass Sleeve Gastrectomy HR (95% CI): 1.58 (1.44–1.74)RD (95% CI): 3.63 (2.85–4.41) No. at risk Roux-en-Y gastric bypass 19,529 19,261 19,040 18,879 18,705 18,546 Sleeve Gastrectomy 19,529 19,404 19,289 19,185 19,088 18,954
Hyperlipidemia
Cumulative incidence, % 0 20 40 60 0 1 2 3 4 5 Time, years Roux-en-Y gastric bypass Sleeve Gastrectomy HR (95% CI): 0.82 (0.79–0.84)RD (95% CI): −7.99 (−10.21 to −5.77) No. at risk Roux-en-Y gastric bypass 19,529 16,019 14,800 14,023 13,380 12,756 Sleeve Gastrectomy 19,529 15,415 14,036 13,082 12,320 11,571
Osteoporosis
Cumulative incidence, % 0 1 2 3 0 1 2 3 4 5 Time, years Roux-en-Y gastric bypass Sleeve Gastrectomy HR (95% CI): 0.99 (0.85–1.14)RD (95% CI): −0.02 (−0.23 to 0.20) No. at risk Roux-en-Y gastric bypass 19,529 19,498 19,454 19,420 19,367 19,310 Sleeve Gastrectomy 19,529 19,489 19,448 19,407 19,356 19,300
Ascites
Cumulative incidence, % 0 1 2 3 0 1 2 3 4 5 Time, years Roux-en-Y gastric bypass Sleeve Gastrectomy HR (95% CI): 2.08 (1.82–2.38)RD (95% CI): 3.11 (2.54–3.68) No. at risk Roux-en-Y gastric bypass 19,529 19,371 19,231 19,140 19,034 18,940 Sleeve Gastrectomy 19,529 19,473 19,409 19,369 19,325 19,266
Figure 1. Study selection flow diagram
Roux-en-Y gastric bypass
Adults with obesity (BMI ≥ 35 kg/m²) identified in TriNetXn = 312,487
After eligibility criteria (age ≥ 18, BMI threshold, no prior bariatric surgery)n = 87,204
After exposure definition (CPT 43644, 43645)n = 19,665
After applying exclusions (T2D, CKD, prior cancer)n = 19,665
Sleeve Gastrectomy
Adults with obesity (BMI ≥ 35 kg/m²) identified in TriNetXn = 312,487
After eligibility criteria (age ≥ 18, BMI threshold, no prior bariatric surgery)n = 87,204
After exposure definition (CPT 43775)n = 56,015
After applying exclusions (T2D, CKD, prior cancer)n = 56,015
Propensity-score matched 1:1 on 90 baseline covariates
n = 19,529 each arm
Supplementary Document
Methods
Data source & cohort
We conducted a retrospective cohort study on the TriNetX Research Network (~140M patients across ~95 healthcare organizations, primarily in the US). Adults with obesity (BMI ≥ 35 kg/m²) who underwent Roux-en-Y gastric bypass (CPT 43644, 43645) or sleeve gastrectomy (CPT 43775) between 2010 and 2024 were identified. Propensity-score matching (1:1, nearest-neighbor, caliper 0.1σ) balanced the cohorts on 90 baseline covariates including demographics, BMI, comorbidities, labs, and medications.
Supplementary Table S3
Eligibility criteria and exposure definitions
CriterionTypeCode / source
Eligibility criteria
Age ≥ 18 years at indexBinaryDemographics: AI
BMI ≥ 35 kg/m²ContinuousLaboratory: LOINC 39156-5
No prior bariatric surgeryBinaryCPT: 43644, 43645, 43775, 43770
Exposure Definition
Roux-en-Y gastric bypassBinaryCPT: 43644, 43645
Sleeve GastrectomyBinaryCPT: 43775
List of abbreviations
BMI - body mass index
CI - confidence interval
CKD - chronic kidney disease
CPT - Current Procedural Terminology
HR - hazard ratio
ICD-10 - International Classification of Diseases, 10th revision
KM - Kaplan–Meier
LOINC - Logical Observation Identifiers Names and Codes
NNT - number needed to treat
PSM - propensity-score matching
RD - risk difference
SMD - standardized mean difference
T2D - type 2 diabetes
TNX - TriNetX
Fully local. No data leaves your browser.
Zero install. Open the page and drop files.
Word, PNG & SVG. Paste straight into a manuscript.
Weeks to minutes
40+
hours reclaimed per manuscript

Conservative estimate of what it takes to hand-build a Table 1, forest plot, Kaplan–Meier panel set, PRISMA flow diagram, and Supplementary Document from a TriNetX export. OutcomePlot gets you there in minutes, not weeks.

5
Publication artifacts

Table 1, forest plot, Kaplan–Meier, PRISMA flow, and Supplementary Document - all in a single workflow.

0
Uploads

Every byte stays in your browser. No account, no server.

<5 min
From drop to export

Open the page, drop files, style, download. Done.

What OutcomePlot makes

Tables, figures, and a full supplement. One workflow.

Everything a journal asks for, generated in one browser session from a single TriNetX export. No axes to tune, no templates to wrestle with, no weeks of manuscript work.

Table 1

Before and after propensity-score matching, with units, n (%), and standardized mean differences. Drag to reorder, click to rename, export to Word.

Forest plot

Hazard ratios, risk differences, E-values, and NNT / NNH on an auto-scaled log axis with a pinned reference line.

Kaplan–Meier

Per-outcome cumulative incidence curves with shaded CI bands, at-risk tables, and HR / RD callouts.

PRISMA flow

Editable flow diagram from your Query Criteria tree, side-by-side per cohort, with automatic n counts and a propensity-score matching terminus.

Supplementary Document

Methods, eligibility and exposure tables (S3), exclusion codes (S4), baseline variables (S5), and a scanner-generated list of abbreviations. Exports straight to Word.

How it works

From export to figure in three steps.

No upload, no install, no account. Open the page and you are ready.

1

Drop your files

The TriNetX Compare Outcomes .docx report, plus any KM CSV exports. Everything parses locally in your browser.

2

Review and style

Rename cohorts, pick colors, reorder outcomes, flag rows to include. Every change previews live.

3

Export

Word for Table 1. SVG and PNG for every figure. Ready to paste into your manuscript, fonts and all.

Built by practicing researchers.

OutcomePlot was built while writing real papers. These peer-reviewed studies used earlier versions of the tool to turn TriNetX exports into journal-ready figures.

Diabetes, Obesity and Metabolism

Comparative effectiveness of SGLT2 inhibitors and GLP-1 receptor agonists for incident dementia: a retrospective multicohort study

Anagnostakis F, Kokkorakis M, et al.2026
Read the article →
Alzheimer’s & Dementia

Impact of cardiometabolic conditions on the progression from mild cognitive impairment to dementia: a large cohort study

Anagnostakis F, Kokkorakis M, et al.2025
Read the article →
Endocrine Practice

Long-term risk of prediabetes and type 2 diabetes following SARS-CoV-2 infection: a nationwide cohort study

Anagnostakis F, Kokkorakis M.2026
Read the article →
Diabetes Research and Clinical Practice

Association of glucagon-like peptide-1 receptor agonist use with anxiety disorders, depression, self-harm, and suicidality: a large cohort study

Anagnostakis F, Kokkorakis M, et al.2026
Read the article →

About OutcomePlot.

We're building tools to accelerate scientific discovery. We help you skip the time spent learning data science and focus on your research. We'll take care of the rest.

OutcomePlot was built to close the gap between a TriNetX Compare Outcomes export and a manuscript-ready figure set. The tool parses the .docx report and KM CSV files entirely in the browser, renders each figure at publication scale, and exports to the formats journals actually want.

Michail Kokkorakis, MD, PhD
Author
Postdoctoral Fellow in the Section of Digestive Diseases at Yale School of Medicine. His research draws on large population cohorts (UK Biobank, Lifelines, TriNetX) and machine learning to study risk factors, patient stratification, and precision strategies for metabolic and alcohol-associated liver disease.
Filippos Anagnostakis, MD
Author
Postdoctoral Researcher at the AI for Brain Imaging and Image Analysis Lab (AIBIL), University of Pennsylvania. He applies machine learning to high-dimensional brain MRI from clinical trials and electronic health records to examine how cardiometabolic conditions shape brain aging.
How to cite Kokkorakis M, Anagnostakis F. OutcomePlot: a browser-based tool for TriNetX Compare Outcomes reporting. 2026. Available at https://outcomeplot.com