Fig. 2

The risk bias evaluation of RCTs included in the meta-analysis. The included RCTs had a low risk of bias for random sequence generation, blinding or participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other biases (green bars for corresponding items). For allocation concealment, 3 RCTs did not report how allocation was concealed, and one did not conceal. Therefore, 25% of included RCTs had a high risk of bias, and 75% had an unclear risk of bias (a yellow and red bar). For blinding of participants and personnel, 1 RCT did not report if blinding was conducted, and 1 was open-label. Therefore, 25% of included RCTs had a high risk of bias, 25% had an unclear risk of bias, and 50% had a low risk of bias (a green, yellow, and red bar)