(This blog is based on the Open Access publication: Paul Glasziou, Mark Jones, Mike Clarke. Setting New Research in the context of Previous Research: some options, BMJ EBM, 2023,)
Interpreting reports of new controlled trials is greatly aided if the Discussion sets the results in the context of the results from available similar research, ideally a meta-analysis. In the paper above, we set out 3 options for doing this:
(1) Update the previous review, that is conduct a new search, extraction and meta-analysis;
(2) Add to the “forest plot” of the existing review, or
(3) Add the new result to the summary effect of the previous meta-analysis.
That last option is the simplest. We suggested including in the trial’s Discussion section a statement such as:
A recent systematic review found that [RESULTS]; adding the results of our new trial to this meta-analysis would change the effect to [NEW RESULTS].
For this 3rd (and simplest) option, the online calculator (link below) provides an updated meta-analysis summary estimate and also a simple forest plot for the Discussion section of your RCT report. The following steps outline how to do this.
- Choose the appropriate Tab from the Excel sheet out of 4 options based on (i) the type of outcome (Relative Risk or Mean Difference) and (ii) type of model (fixed or random, with fixed as the default).
- Input the relevant data (6 numbers) into the Table below: the previous meta-analysis summary result and new trial result; risk ratio and 95% confidence interval for dichotomous outcomes and mean difference and 95% confidence interval for continuous outcomes.
- [Optional] If you choose the random effects option, then you to also need to input the value of tau2 reported from the previous random effects meta-analysis. Using the previous value for tau2 will provide new estimates based on the previous level of heterogeneity hence it is an approximation only. A more accurate result will be obtained by conducting an updated meta-analysis by incorporating the new trial result to the previous meta-analysis of individual trial results.
Using the Calculator
Step 1: Choose your outcome type (click one of the 4 tabs in the spreadsheet):
- Relative Risk – fixed effect
- Relative Risk – random effect
- (Standardized) Mean Difference – fixed effect
- (Standardized) Mean Difference – random effect
Step 2: Fill in the 6 empty cells in the Table below (either using the online spreadsheet, or download the spreadsheet below).
| Estimated effect | Lower CI | Upper CI | |
| Previous Meta-Analysis | |||
| New Trial |
(CALCULATE will produce the plot and numbers below)
You can directly enter into the web version below (or download the xls and use offline).
You can also download the Excel Spreadsheet from this link (you will need to download to edit the spreadsheet). This spreadsheet will also provide a simple “forest plot”, but this is provided for use in talks etc, not necessarily for inclusion in the publication describing the trial.
Please feel free to download and disseminate to colleagues or journal editors. Feedback and suggestions very welcome.
Healthy Evidence 