In proportion/prevalence meta-analysis how to test the pooled prevalence of  two or more subgroups are statistically different using Stata?

More Deepani Siriwardhana's questions See All
Similar questions and discussions