I am performing a meta-analysis of the correlation of a HLA DRB1 with RA.
In similar research different programs were used in Data analysis (calculation of pooled Odds ratio, CI95%, p-value, Q, I2 etc)
The two most commonly used are:
Revman ( Review Manager | Cochrane Comunity)
CMA (Comprehensive Meta-Analysis)
Different results were given when I applied the meta-analysis in both programs.
RevMan calculates pooled Odds Ratio by using Cochran-Mantel-Haenszel method, when CMA calculates pooled Odds Ratio by using Borenstein et al method.
What software (statistical method) is most recommended?
Is there any statistical significant difference between the results of the two methods?
Thank you