Fig. 2

a-b. Risk of Bias Assessment Using Cochrane the Collaboration’s Tool. The included studies were evaluated for methodological flaws using the Cochrane Collaboration’s risk of bias assessment tool. [25] a illustrates the risk of bias for each study. Studies were indexed by the last name of the first author and year of publication. Seven domains of risk of bias were assessed for each study, including random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and industry funded. , , denote low, unclear, and high risk of bias, respectively. Six of the 12 studies were assessed as having high risk of attrition bias and two studies were funded by industry (high risk of bias). Ten of the 12 included studies did not report random sequence generation and allocation concealment (unclear risk of selection bias). In contrast, the blinding of participants and personnel were well performed (low risk of performance bias in seven of the 12 included studies). Selective reporting was not observed in any studies (low risk if bias). Figure 2b shows the overall risk of bias by domain: the risk of bias is displayed as low risk (green, +), unclear (yellow,?), or high risk (red, −)