Please could someone recommend a resource that explains performing meta-regression with multiple moderators. I have read the STATA META ANALYSIS REFERENCE MANUAL, but this only gives examples of entering one variable as a moderator.

I would ask on statslist.org but the website has been down for a number of days now. If someone knows an alternative online forum, please let me know.

I understand how to run the multiple meta-regression, but given the coefficients, p values, and residual heterogeneity all change for the same moderator when run as a single moderator compared to when run as one of several moderators, I would like to understand what is going on and how to interpret the output properly for reporting in my meta-analysis.

Finally, authors often sometimes talk about being able to find the "model of moderators that best explain the heterogeneity" - is there any way to ask STATA to find this model, by entering all the covariates as moderators. I.e. the output will say which combination of moderators is the best fit model?

Many thanks.