Forest plot
- Concept
A forest plot shows each study as a square with its interval, and the pooled effect as a diamond, against the line of no effect.
What it is
The forest plot summarises a meta-analysis. Each row is a study: the square marks its effect and its size shows the weight; the horizontal line is its confidence interval. The diamond at the bottom is the pooled effect: its centre is the estimate and its width the CI. The vertical line of no effect sits at 1 for ratios (RR, OR, HR) and at 0 for differences.
Worked example
- Data (made up): 4 trials of a falls-prevention programme; RR (95 % CI): 0.70 (0.50–0.98); 0.85 (0.60–1.20); 0.60 (0.40–0.90); 1.05 (0.70–1.58).
- Trials 1 and 3: their CI lies entirely below 1: a significant reduction in each.
- Trials 2 and 4: their CI crosses 1: on their own they show no difference.
- Inverse-variance weights on the log scale: ≈ 30 %, 28 %, 21 % and 21 %: the first two, with narrower CIs, weigh more.
- Diamond (fixed effect): RR 0.78 (95 % CI 0.65 to 0.94); it does not touch the line at 1.
- Relative reduction: 1 − 0.78 = 0.22, that is, 22 % fewer falls.
- Reading: two inconclusive trials do not cancel the set; combined, the 4 show a benefit with a CI that excludes 1.
Markers
- Square
- CI
- Diamond
- Line of no effect
- I²
In practice
- Look first at the diamond and the line of no effect, then at the spread of the squares: if they line up, the result is consistent.
- A narrow diamond close to 1 can be significant yet clinically small: translate it into an absolute effect.
Common pitfalls
- Vote counting (2 positive and 2 negative, so we do not know): meta-analysis weighs and combines, it does not count.
- Reading the wrong side of the axis as favouring the intervention.
- Forgetting that the ratio axis is logarithmic: 0.5 and 2 are the same distance from 1.
Test yourself
Think of the answer, then tap the question to see it.
What does the size of each study's square in a forest plot show?
The study's weight in the meta-analysis, which depends mainly on its precision (sample size and number of events).
Where is the line of no effect for a relative risk and for a mean difference?
At 1 for the relative risk (and other ratios) and at 0 for a mean difference.
The diamond of an RR forest plot runs from 0.65 to 0.94. What does it mean?
That the pooled effect is a statistically significant reduction: the whole CI lies below 1.
Connections
Educational content on epidemiological and statistical methods; the examples use made-up numbers to teach the calculation. Not a clinical or legal guide: current Colombian rules are in the official sources.
Sources: Chapter 10: Analysing data and undertaking meta-analyses (Cochrane, 2026).
Spotted a mistake?We are always improving for the community. Tell us what is wrong and our team will review it.Report a mistake
Representative model, for educational purposes only.