Publication-Ready Logistic Regression Tables using gtsummary

A workshop write-up on building publication-ready binary logistic regression tables in R with gtsummary.

This free workshop, hosted by Unijos R Users — a non-profit, volunteer-run community — walked participants through building publication-ready statistical tables in R using the gtsummary package.

Session Overview

Date: Saturday, January 17, 2026 Format: Online (Zoom) Attendance: 13 participants

Topics Covered

Binary Logistic Regression Fundamentals

• Understanding odds ratios and their interpretation
• When to use logistic regression vs other methods
• Model assumptions and diagnostics

Variable Selection Strategies

• A priori selection based on literature
• Backward elimination approach
• Avoiding overfitting and stepwise selection pitfalls
• Sample size considerations (Events Per Variable rule)

Creating Professional Tables with gtsummary

• Univariate and multivariable regression tables
• Customizing table appearance and formatting
• Adding model performance statistics
• Merging multiple analysis results
• Exporting to Word and other formats

Model Assessment

• Hosmer-Lemeshow goodness-of-fit test
• ROC curves and C-statistic interpretation
• Model discrimination and calibration

A Minimal Example

library(gtsummary)

# Fit a logistic regression model
model <- glm(response ~ predictor1 + predictor2,
             data = mydata, family = binomial)

# Publication-ready table
tbl_regression(model, exponentiate = TRUE)

Want Materials or a Recording?

As a non-profit community, we share our workshop materials freely with members. Reach out at info@unijosrusers.org or join our Meetup group for upcoming sessions and resources.