OR meta-analysis using R

[Esc]

Hi!

I have currently been using MedCalc for Odds Ratio meta-analysis. For both intervention and control groups is required the following data input: number of positive cases and number of total cases.

The problem is, sometimes studies report Odds Ratios with CI:s, without primary data. ORs without primary data cannot be entered to a pooled analysis in MedCalc.

I wonder if both data types (ORs with CI and primary data) could be combined at same time using R? Could you give any examples how to input such data.

Thank you!

 

[hlsmith]

I would look at the "metafor" package in R. I haven't seen direct input of ORs in it, but that doesn't mean it is not an option. Another, way less ideal approach could be to convert things to Cohen effect sizes. I am guessing either will require you to calculate the SEs via backtransforming the confidence intervals.