Table 1.

Risk of bias from previous meta-analyses.

Study name	Linsenmeyer	Waites	Lee	Hess
Year	2004	2004	2007	2008
	Random sequence generation
Jepson et al. 2012 [17]	Unclear	Unclear	Low	Unclear
Wang et al. 2012 [18]	Not reported	Unclear	Not reported	Unclear
Luis et al. 2017 [19]	Not reported	Unclear	Low	Unclear
	Allocation concealment
Jepson et al. 2012 [17]	Unclear	Unclear	Low	Low
Wang et al. 2012 [18]	Not reported	Unclear	Not reported	Unclear
Luis et al. 2017	Not reported	Low	Low	Unclear
	Blinding of participants and personnel
Jepson et al. 2012 [17]	Low	Low	Low	Low
Wang et al. 2012 [18]	Not reported	Low	Not reported	Low
Luis et al. 2017 [19]	Not reported	Low	Low	Low
	Blinding of outcome assessment
Jepson et al. 2012 [17]	Low	Unclear	Low	Low
Wang et al. 2012 [18]	Not reported	Not reported	Not reported	Not reported
Luis et al. 2017 [19]	Not reported	Low	Low	Low
	Incomplete outcome data
Jepson et al. 2012 [17]	Low	High	Low	High
Wang et al. 2012 [18]	Not reported	High	Not reported	Unclear
Luis et al. 2017 [19]	Not reported	Low	Unclear	Unclear
	Selective reporting
Jepson et al. 2012 [17]	Low	Low	Low	Low
Wang et al. 2012 [18]	Not reported	High	Not reported	High
Luis et al. 2017 [19]	Not reported	Low	Unclear	Low
	Other bias
Jepson et al. 2012 [17]	Unclear	Unclear	Low	Low
Wang et al. 2012 [18]	Not reported	Not reported	Not reported	Not reported
Luis et al. 2017 [19]	Not reported	Low	Low	Low