Consider this example:
Some studies report the effect of treatment X at month 1 after intervention.
Some studies report this effect for the 2nd month after the initiation of intervention.
Some others report it for the 3rd month.
Now I want to preform a meta-analysis for studies at each month, then do a regression analysis on the resulted summary effects to evaluate its change during time.
This time of treatments is a variables which is varying in all papers and can thus be taken as a moderator.
pls also refer to https://www.youtube.com/channel/UCj38Topqj8yC3SdrV6HypIg
The studies to be analyzed will be clinical trials, randomized or not, with or without a control group. It has to be defined by the inclusion / exclusion criteria. You should do a separate meta-analysis for each moment in time, with its corresponding forest plot.
I agree with Elisabet Roca -Millan. The design of the studies can be a Randomized Clinical Trial with/without control depending on your inclusion criteria and the analysis need to be separate for the reported time given the effect could vary depending on the time range.
- Badges
- Science method
- Similar topics
- Methodology