We want to conduct meta-analysis on papers that report a % change in vaccination coverages. But I am not sure if at all a metanalysis is possible using % change.
If you mean relative % change eg increase from 50% coverage to 60% coverage is 10/50 = 20% increase then you could just calculate the mean of these % increases to get an overall effect measure.
If you mean absolute differences in % vaccinated in populations, eg in above example difference in % vaccinated = +10% ,to get an overall effect measure you would need population sizes .