Reading a forest plot
A forest plot is the diagram of a meta-analysis, which combines the results of several studies, with one square for each study and a diamond for the combined result.
Remember
The diamond is the combined result: its centre is the combined odds ratio (the odds of an event in one group divided by the odds in the other) and its width is the combined confidence interval (the range of values compatible with the result).
Memory aid
The diamond is the bottom line: the whole meta-analysis in one shape, sitting at the bottom of the plot.
Look to the bottom of the plot, where the diamond usually sits, for the whole meta-analysis in one shape.
Each study in a meta-analysis appears as a square at its odds ratio, on a shared scale. A horizontal line through the square is that study's confidence interval.
The size of the square shows the study's weight, which is how much influence it has on the combined result. In general, a narrower confidence interval and a larger sample give more weight and so a bigger square. A long horizontal line therefore normally goes with a small square, because a wider interval means a less precise estimate.
The vertical line of no effect sits at an odds ratio of 1, where the odds are equal in the two groups. It is a fixed comparison point, not a result. If a study's confidence interval crosses it, there is no clear difference. A study clearly favours one group only when its confidence interval stays entirely on one side of that line. Some plots use a mean difference instead of an odds ratio, and the line of no effect is then at 0, meaning there is no difference between the group means.
The diamond, usually shown at the bottom of the plot, is the pooled (combined) result of the meta-analysis. Its centre, often marked by a dotted line, is the pooled odds ratio, and its width is the pooled confidence interval. If the diamond's confidence interval crosses the line of no effect, no clear difference has been shown overall.
Key facts
- Square
- marks one study's odds ratio, and its size shows that study's weight.
- Weight
- is usually higher for a study with a larger sample and a narrower confidence interval, which gives it a bigger square.
- Horizontal line
- is the confidence interval of that one study, so a longer line means a less precise study.
- Line of no effect
- sits at an odds ratio of 1, where the odds are equal in the two groups.
- Centre of the diamond
- is the pooled odds ratio.
- Width of the diamond
- is the pooled confidence interval.
- Vertical position of a study
- is just the order in which the studies are listed, so it does not measure weight.
Exam traps
- The size of a square is not the pooled result, because it shows the weight of that one study.
- The vertical line at 1 is not the pooled effect, because the pooled effect is shown by the diamond.
- A longer horizontal line does not mean more weight, because it shows a wider confidence interval and a less precise estimate, and such a study normally gets less weight.
Read the full notes
Tested in 7 SimpleMed+ questions
Practice it in Population and Social Science, with the answer and a full explanation after every question.
Practice these questions