1. Create an analysis group
To set up an analysis group (formerly called 'comparison'), click on the '+Add Analysis Group' button in the 'Analyses' section.
2. Name the analysis group
Click on the 'Edit Analysis Group' button and enter the name of the new analysis group. Note: you cannot set labels for the intervention groups and graph labels until you add an analysis for this analysis group. After you add an analysis (previously called 'outcome'), go back to the 'Analyses' section, open the newly created analysis group, click on the 'Edit Analysis Group' link and set the graph and group labels for all analyses under this analysis group. Note: you need to click on the 'Set Labels' buttons to save the label names.
Click on the 'Back to Analyses' link to return to the list of comparisons.
3. Create an analysis
Click on the plus icon to the right-hand side of an analysis group to create a new analysis. You can also do that by clicking on the '+Add Analysis' button when you are in the analysis group.
4. Name the analysis
Type in the name of the analysis. (Now you can go back and set the labels for the intervention groups and graph labels – see point 2 above.)
5. Set the analysis options
Click on the 'Options' tab to choose the options for meta-analysis and display of results. Note: you can only change the data type (e.g. dichotomous to continuous) while you are editing the analysis for the first time. When you close the analysis, you cannot change the data type.
Under the ‘Options’ tab you can also select the statistical method, effect measure, analysis model, and how you want to display the results. If you are doing a subgroup analysis, the test for subgroup differences is checked by default. You can swap event and non-event and change the confidence intervals.
If the outcome in this Analysis was pre-specified in the protocol for a Risk of Bias 2 assessment, check the ‘Show the risk of bias table’ box to create a risk of bias table for this Analysis that includes all domain judgements and support for judgements for this outcome. Note that this option is only available if Risk of Bias 2 is enabled in your review.
6. Add studies and study data
Once you have selected the analysis options, go back to the 'Data' tab and click on the '+Add Data Row' button to add a study from the list of included studies.
You can then enter the data into the table either by typing the data manually, or by copying and pasting data (in tabular form) from an external source.
You can delete a data row by clicking on the 'Action' button to the right and selecting 'Delete Data Row'.
Note, it is also possible to use the button Copy analysis to clipboard to copy the data in an analysis in RevMan and paste it in an external file (such as in a text editor or Excel file). The headings and any confidence interval ranges are not copied and would need to be added manually.
See Calculator to transform raw data extracted from study reports when it does not match the format required for meta-analysis.
7. Adjust how the forest plot is displayed
Once you have entered the data for the studies included in the analysis, click on the 'Graphs' tab to adjust the forest plot. By default, the scale is set to 100 and you can change that to optimize the presentation of your results.
You can select to include the risk of bias summary in the forest plot by selecting 'Show Risk of Bias summary'.
You can sort the studies by study ID, year of study, weight, effect size, or risk of bias item.
All settings in the Graph tab will be reflected in the published review in the forest plots presented in the associated Analysis section.
8. Add a footnote to a study in forest plot
You can add a footnote to a particular study in an analysis and forest plot by editing the Analysis, clicking on Data tab, selecting a study row, and typing in a footnote in the box below the analyses. This can be done for as many studies as necessary, and each footnote will be numbered.
These footnotes will then appear below the Forest plot figure. You can see this by clicking on the Graphs tab.
Please note, long footnotes could be cut off at the end. Please keep footnotes as concise as possible, and check in the View of the current version of the review to ensure long footnotes are not cut off.