Forest Plot
A chart used in meta-analysis to display effect sizes and confidence intervals from multiple studies — each study is a row with a point estimate and a horizontal CI line, anchored by a vertical line of no effect and a diamond summary at the bottom.
// 01 — The chart
What it looks like
A forest plot showing five studies and a pooled summary diamond. Each row displays a study’s point estimate (square) and 95% confidence interval (horizontal line). The dashed vertical line marks no effect (OR=1).
// 02 — Definition
What is a forest plot?
A forest plot (also called a blobbogram) is the standard way to display the results of a meta-analysis — a statistical technique that combines findings from multiple independent studies addressing the same question. Instead of reading a table of numbers, you get a visual summary that makes it easy to see whether individual studies agree and what the overall evidence says.
Each study appears as a horizontal row. The point estimate (usually an odds ratio, risk ratio, or mean difference) is shown as a square whose size is proportional to the study’s weight in the analysis. The horizontal line through the square represents the confidence interval — the range within which the true effect plausibly lies.
A vertical reference line (the “line of no effect”) divides the plot. If a study’s confidence interval crosses this line, the result is not statistically significant. At the bottom, a diamond shows the pooled estimate — the combined result of all studies. The diamond’s width represents its confidence interval.
Origin: The forest plot was popularized in the early 1990s by the Cochrane Collaboration (now Cochrane) for use in systematic reviews of healthcare interventions. The name “forest” comes from the forest of lines that the plot resembles when many studies are included.
// 03 — Anatomy
Parts of a forest plot
// 04 — Usage
When to use it — and when not to
- Presenting results of a meta-analysis that combines multiple studies
- You need to show both individual study results and the pooled summary
- Comparing effect sizes and their precision across studies
- Communicating heterogeneity — how much studies agree or disagree
- Visualizing subgroup analyses within a systematic review
- Your audience is familiar with confidence intervals and effect sizes
- You have only a single study — a simple table or bar chart suffices
- Your data isn’t from a meta-analysis or systematic review
- Your audience lacks statistical literacy (use simplified visuals instead)
- You need to show trends over time — use a line chart
- Studies measure fundamentally different outcomes that shouldn’t be pooled
- There are so many studies (50+) that the plot becomes unreadable
// 05 — Reading guide
How to read a forest plot
Follow these steps whenever you encounter a forest plot in the wild.
Find the line of no effect
This vertical dashed line is your anchor. For odds ratios and risk ratios it’s at 1; for mean differences it’s at 0. Everything to the left favours one group, everything to the right favours the other.
Read each study row
Each row shows one study. The square is the point estimate (the study’s best guess of the effect). The horizontal line is the 95% confidence interval. A wider line means more uncertainty; a larger square means the study carries more weight.
Check which side of the line studies fall on
If most squares and their CIs sit on the same side of the no-effect line, the evidence consistently points in one direction. If they scatter on both sides, the evidence is mixed.
Look for the summary diamond
The diamond at the bottom represents the pooled result. Its center is the overall effect estimate, and its width is the confidence interval. If the diamond doesn’t cross the no-effect line, the pooled result is statistically significant.
Assess heterogeneity
Look at how much the individual CIs overlap. Poor overlap (wide scatter of point estimates) suggests high heterogeneity — the studies may not be measuring the same thing. Check the I² statistic if reported.
// 06 — Pitfalls
Common mistakes
Not scaling squares by study weight
Fix: Square size should reflect each study’s weight (typically inverse variance). Equal-sized squares mislead readers into thinking all studies contribute equally to the pooled result.
Using a linear scale for ratio measures
Fix: Odds ratios and risk ratios should be plotted on a logarithmic scale so that effects of equal magnitude appear symmetrically around the line of no effect.
Omitting the summary diamond
Fix: The diamond is the whole point of a meta-analysis forest plot. Without it, readers see individual studies but miss the combined conclusion.
Pooling fundamentally different studies
Fix: If studies measure different outcomes, populations, or interventions, pooling is misleading. Use subgroup analyses or acknowledge that pooling is inappropriate.
Ignoring heterogeneity
Fix: Always report and discuss I² and Q statistics. High heterogeneity (I² > 75%) means the pooled estimate may not be meaningful. Consider a random-effects model or subgroup analysis.
// 07 — In the wild
Real-world examples
Cochrane systematic reviews
The Cochrane Library publishes thousands of systematic reviews on healthcare interventions, each featuring forest plots as the primary visual summary. Their standard format — with study weights, heterogeneity stats, and subgroup diamonds — has become the de facto template for medical meta-analysis.
COVID-19 vaccine efficacy
During the pandemic, forest plots were widely used to compare vaccine efficacy across trials. Each trial appeared as a row, with the pooled diamond showing overall efficacy. These plots appeared in journals like The Lancet, NEJM, and BMJ, reaching both specialist and general audiences.
Education intervention research
The Campbell Collaboration uses forest plots to synthesize evidence on educational programs. For example, a meta-analysis of tutoring interventions might show effect sizes from 20 studies, revealing whether tutoring consistently improves outcomes across diverse contexts.
// 08 — Quick reference
Key facts
// 09 — Variations
Types of forest plots
Forest plots adapt to different analytical needs while keeping the same core structure.
Standard forest plot
The classic format with study rows, weighted squares, CI lines, and a summary diamond. Used in most Cochrane reviews.
Subgroup forest plot
Groups studies by a characteristic (dose, population, design) with separate sub-diamonds and an overall diamond at the bottom.
Cumulative forest plot
Studies are added one at a time (often chronologically) to show how the pooled estimate evolves as evidence accumulates.
Prediction interval forest plot
Adds a prediction interval band around the summary to show where the true effect in a future study is likely to fall.
// 10 — FAQs
Frequently asked questions
What is a forest plot?+
A forest plot (also called a blobbogram) is the standard way to display the results of a meta-analysis — a statistical technique that combines findings from multiple independent studies addressing the same question. Instead of reading a table of numbers, you get a visual summary that makes it easy to see whether individual studies agree and what the overall evidence says.
When should you use a forest plot?+
Use a forest plot when presenting results of a meta-analysis that combines multiple studies. It also works well when you need to show both individual study results and the pooled summary, and when comparing effect sizes and their precision across studies.
When should you avoid a forest plot?+
Avoid a forest plot when you have only a single study — a simple table or bar chart suffices. It is also a poor fit when your data isn’t from a meta-analysis or systematic review, or when your audience lacks statistical literacy (use simplified visuals instead).
Is a forest plot suitable for dashboards?+
Yes — a forest plot can work well in dashboards as long as the panel is large enough for readers to perceive the encoded values, has a clear title, and includes the legend or axis labels needed to interpret it.
What category of chart is a forest plot?+
Forest Plot belongs to the Statistical family of charts. Charts in that family are designed to answer the same kind of question, so they often work as alternatives when one doesn't quite fit your data.
How do you read a forest plot?+
Start with the axis labels and legend, then look at the overall shape before zooming into individual marks. Compare prominent features against the rest of the data, and verify any conclusion against the underlying numbers when precision matters.