Literature or systematic reviews may be performed to attempt to make sense of the results from all of these studies by comparing and synthesising the findings. However, some reviews go beyond this by providing a new pooled result of the effectiveness of a particular treatment. This pooled result is calculated using the quantitative results from the studies by a method called “meta-analysis”.
We can use a meta-analysis to combine these varied results together to get an overall impression about what the results from all of the studies are saying and to get a single pooled result of how much the treatment is beneficial.
Meta-analysis does not simply involve averaging the results of the individual studies, but requires a statistical method which combines the results whilst taking into account the sizes of the studies.
A common belief is that meta-analyses should only be based on results from clinical trials. However, assuming the proper statistical methods are used; meta-analyses can be based on data from other types of study designs, such as case-control studies and surveys.