Background: I have incidence rates from different surveys, done in different zones of a region. I want to combine the results of several zones so as to make a general statement about the region. I do understand that the denominator of the incidence rate takes into account the sample size and exposure time. 

Confusion: Is weighing each of the estimates ( say using sample weight ) necessary given that the denominator has already taking care of the differences in sample size and exposure period?

Secondly, I want to know, if necessary, how to incorporate sample precision into this process?