Help creating log metric for pooled odds ratio

[michael lee smith]

Dear all,

I am trying to perform egger test on my meta analysis of studies on the prevalence of disease A in people drinking alcohol. A sample is shown below (I do not have 2x2 table data for all the studies)

OR UCI 95% LCI 95%
0.84 0.74 0.93
0.83 0.75 0.92
1.89 1.08 3.32
1.38 0.77 2.48

On entering meta bias egger it says - You must declare your meta-analysis data using either meta set or meta esize.

I have difficulty performing this step. The help guide says:
If you are working with effect sizes in the original metric, such as odds ratios or hazard ratios, with meta set, you should specify the effect sizes and CIs in a normalizing metric, such as the log metric.

However I am not sure how to create the log metric.

Any help would be appreciated!

[Clyde Schechter]

foreach v of varlist or uci lci { // USE YOUR ACTUAL VARIABLE NAMES HERE
    gen `v'_log = log(`v')
}