Risk of bias within studies
Figure 3a displays risk of bias assessment for the efficacy outcome in nine studies. Three had low overall risk of bias, three were high, and three were unclear. Randomization processes were adequate in all studies, but only four followed the intended interventions for the participants. Two studies had high risk of bias due to intervention deviations for failure to report the analysis used to estimate the effect of assignment to intervention and failure to analyze participants in the group to which they were randomized. Three had unclear risk of bias for not reporting blinding or possible deviations.
One study did not report the reason for missing outcomes. Two studies may have used inappropriate outcome measurement methods.  Three studies had unclear risks due to selective reporting and insufficient result detail.
Figure 3b shows bias risk for safety outcomes in eight studies. Three had low overall bias risk, three high, and two were unclear. Unclear and high risks were due to intended intervention deviations, missing outcome data, inappropriate outcome measurement, and selective reporting of results.