How to Create the Plot
Start the program, or use New table/graph to create a Column data table. Keep the default choice to enter replicates into columns (in this case they are not replicates, but Prism treats them as such).
Enter the data. For each group, enter the odds ratio and its high and low confidence limits (computed elsewhere) into one column. Each treatment group goes in a separate column with three values per column. The order of those three values does not matter. Label the groups using column titles.
Click on the Graphs tab.
Choose the thumbnail for your desired graph style. Also choose to plot the median and range: the median of your three values is the odds ratio, and the range of the three values spans the confidence interval.
View the graph.
The default graph is vertical. To make it horizontal (standard for Forest plots), double-click the graph to open Format Graph, then go to the third tab.
Adjust symbol sizes if needed. Some meta-analysis plots use larger symbols for studies with more subjects. You can adjust each symbol's size manually in the Format Graph dialog.
Edit axis and graph titles. Add a custom dotted line at X = 1.0 using Format Axes to mark the line of no effect. This completes the Forest plot.
Using a Logarithmic X Axis
Epidemiologists often prefer a logarithmic X axis for odds ratio plots. This makes sense because it renders odds ratios greater than 1.0 and less than 1.0 symmetrical. For example, an odds ratio of 2.0 becomes visually symmetrical with an odds ratio of 0.5.
To apply log scale in Prism: Format the X axis with a Log 10 scale and set the range from 0.1 to 10 with "antilog" numbering. Add a grid line at X = 1 to mark the line of no effect.
Tips
- This example plots results from three studies, but you can add more by using additional columns (D, E, F, etc.).
- The median of the three values you enter per column is the odds ratio; the range spans the confidence interval.
- Adjust symbol size manually for studies with different numbers of subjects to reflect study weight.
- Keywords to search: horizontal error bars, odds ratio, meta-analysis, Forest plot.