We considered the domains for each study and then made a judgment of high or low risk of bias for the studies included in the meta-analysis.