pubmed.info
Drag tiles by their title, resize at the bottom edge.

Forest plot generator – meta-analysis online

Enter the results of the individual studies and get the pooled effect, heterogeneity statistics and a publication-ready forest plot. Free, no account; your data are not uploaded, everything is calculated in your browser.

More about this tool

What you can do

  • Pool 2 to 200 studies with the inverse-variance method: fixed effect, random effects or both models side by side.
  • Estimate Tau² with DerSimonian-Laird or Paule-Mandel and optionally apply the Hartung-Knapp adjustment, recommended with few studies.
  • Get Cochran’s Q, Tau², I² with confidence interval and, from three studies, a prediction interval for random effects.
  • Check funnel plot asymmetry with Egger’s test (meaningful from about ten studies).
  • Draw the forest plot with squares sized by weight, confidence intervals and diamonds, optional weight and count columns, heterogeneity below the chart, your own axis range and labels such as “favours treatment”.

How it works

  • Events and sample sizes: one row per study with name, events and total in the treatment group, then events and total in the control group; effect as odds ratio, risk ratio or risk difference. For OR and RR, 0.5 is added to all cells of a study with a zero cell.
  • Means: name, then mean, SD and sample size of the treatment group and of the control group; effect as mean difference or standardised mean difference (Hedges g).
  • Effect estimates: name, effect and the limits of its confidence interval, e.g. hazard ratios; the standard error is derived from the interval, for ratios on the log scale.
  • Separate the values by tab or semicolon, e.g. pasted from Excel. Ratio measures are shown on a logarithmic axis.

Export

SVG with real text and your input embedded, vector PDF (Latin script only) or PNG at 300 or 600 dpi in common journal column widths, plus the results as CSV and a suggested text for methods and results. Load the SVG or a saved project file to continue later.

Your data: nothing is transmitted to a server. Only this browser remembers your input and settings.

Frequently asked questions

Fixed effect or random effects?

The fixed-effect model assumes one common true effect; the random-effects model lets the true effect vary between studies and gives small studies relatively more weight. Choose the model before the analysis on subject-matter grounds, not according to the result.

How do I interpret I²?

I² is the proportion of total variability that is not due to chance; as a rough guide, from about 50% heterogeneity is called substantial and from 75% considerable. With few studies Tau² and I² are very uncertain.

What is not included?

Meta-regression, subgroup and network meta-analyses, Peto and Mantel-Haenszel methods and risk-of-bias assessment. For the study selection, use the PRISMA 2020 flow diagram on pubmed.info.

pubmed.info sets no tracking cookies and builds no user profiles. Visits are counted anonymously (cookieless, shortened IP addresses). Ads from Google appear only according to your consent choice.