A complete application walkthrough with practical instructions for setting up studies, generating forest plots, and exporting professional outputs.
Introduction
The Forest Plot Generator is designed for researchers and clinicians who want to perform transparent meta-analysis and produce publication-ready forest plots without manual calculations. It supports continuous outcomes using Cohen's d and binary outcomes using risk ratio (RR) or odds ratio (OR).
Step 1: Choose your analysis
- Outcome Type: Select Continuous for Cohen's d or Binary for RR/OR.
- Effect Measure: For binary data, choose between Risk Ratio and Odds Ratio.
- Model: Choose Fixed Effect for inverse-variance weighting, or Random Effects for between-study heterogeneity.
- Confidence Level: Select 95%, 99%, or 90% confidence intervals.
Step 2: Enter your study data
Enter each study as a separate row. The required fields depend on the selected outcome type:
- Continuous: group sizes, means, and standard deviations for both arms. Provide either a pooled SD or both group SDs.
- Binary: event counts and total participants in each treatment arm.
Use Add Study to insert additional rows. Use Load Example to populate sample data, and Clear All to reset the form.
Step 3: Generate results
Click Generate Forest Plot after entering at least two studies. The app will validate your data, highlight any issues, and then compute the effect sizes, confidence intervals, heterogeneity statistics, and forest plot.
Understanding the output
- Summary panel shows pooled effect size, confidence interval, number of studies, I² heterogeneity, τ² variance, and z-test p-value.
- Forest plot displays study estimates, confidence intervals, weighting, and the pooled diamond estimate.
- Detailed tables provide per-study effect sizes, original-unit conversions, and fixed/random weights.
Export options
Use the Download Guide as PDF button to save this guide as an elegant professional document. Use the plot download button on the results panel to export the forest plot as a PNG image.
Best practices
- Double-check input values for each study before generating results.
- Use the random-effects model when heterogeneity is substantial (I² > 50%).
- Review the interpretation panel to understand whether the confidence interval crosses the null value.
Prepared by
Manjunath Kulkarni
drmanjunath.in