Statistical analysis
We examined the heterogeneity assumption using Cochran’s Q test and
quantified it using Higgins I2 statistic, withI2 > 50% considered statistically
heterogeneous. We used bivariate mixed-effects regression model to
calculate
pooled sensitivity and 95% confidence intervals (CI). We performed
meta-regression and subgroup analysis to investigate the potential
impact of variates on sensitivity and the source of heterogeneity.
Results were displayed graphically on forest plots. We conducted
sensitivity analysis to evaluate the robustness of the results by
removing studies with high weight or debatable methodology, modifying
stratified criteria, or altering statistical model. We did all
statistical analyses with STATA (version 15.0; Stata Corporation,
College Station, TX, USA)