Statistical Analysis
Meta-analysis of prevalence estimations was done using the Stata commandmetaprop, that allows computation of 95% confidence intervals
using the score statistic and the exact binomial method and incorporates
the Freeman-Tukey double arcsine transformation of proportions to
stabilize the variance and to achieve approximate normality [20].
Meta-analysis of effect sizes that account for the association between
age groups and severity of the disease were done by means of
random-effects models and the method used for estimating between-study
variability was restricted maximum likelihood. Additionally, using
I2 index was obtained to measure heterogeneity.