To comparing between 2 variables from an case-control studies, In the meta-analysis, what is the most appropriate effect measure should be used to estimate a risk of outcomes variables ?
Hi,
Not sure if I understood your query correctly.
1. If the outcome is a categorical variable and the all the studies included in the systematic review are case-control studies, you have to use the odds ratios (OR). Ideally, you should use the adjusted OR (provided in the individual studies) using the generic inverse variance method. Using adjusted OR tries to address some of the biases associated with case-control studies.
2. If the outcome is a continuous variable, you can use the mean and SD of the two groups (cases and controls). Again, it is preferable to use the adjusted mean difference and 95% CI if available. Usually, linear regression is used to adjust for different confounders.
I hope this helps.
Best
Jeeva
Dear Jeeva Sankar ,
Thank you for your answer, but, do you think ,what the best software should I use to perform this analysis?
The easiest software to perform meta-analysis is RevMan (available for download at http://tech.cochrane.org/revman/download).
You can refer to the handbook for help.
Let me know if you need any assistance.
Jeeva
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis