Table 2.

Risk and bias analysis of included studies

Studies	Finkelstein& Cha, 2016 USA	Buller DB, et al. 2014, Australia	Heffner LJ, et al. 2014, USA	Ubhi KH, et al. 2015, England	Zeng et al. 2015, USA	Zeng et al. 2016, USA	Cheung et al 2015
Random sequence generation (selection bias)	High	Low	Low	High	Low	High	Low
Allocation concealment (selection bias)	High	Low	Low	High	Unclear	High	Low
Blinding (performance bias and detection bias) all outcomes	High	Unclear	High	Low	Unclear	Low	High
Incomplete outcome data (attrition bias)	Low	Low	Low	Low	Low	Low	Low
Other bias (analysis bias, publication, )	Low	Low	Low	High	Low	High	Unclear

High – reviewers observed the high level risk and bias

Low- reviewers observed low level of risk and bias

Unclear: Not enough information to judge the criteria