I am conducting a systematic review and using RevMan5 software for meta-analysis of studies. Several studies/references present data for multiple trials - is it possible to input multiple datasets for a single study/reference - if so, how? For example, one paper presents four immunization trials i.e. four datasets in the one study/reference - can I input them all and include all in analysis?
Assuming that all of the reported trials in the study of interest are eligible based on your pre-specified criteria, I think a way to get around this is to label the trials as if they are separate studies (e.g., Trial 1 will be named "Allan 2021a," Trial 2 will be named "Allan 2021b," and so on). However, this will entail separate risk of bias assessments of the trials in the default specifications of RevMan - which I think makes sense because I don't think that a blanket/single risk of bias assessment will give justice to all the trials in the included study.
I came across some meta-analyses that include the same study has multiple datasets. I think this condition is not an accurate way, but no other way is possible in various situations (I could not find). And I included the same study twice in an analysis like "Allan 2021a" and "Allan 2021b".
- Badges
- Science topic