Table 1.

Quality assessment of the included studies

Authors	Sequences generation	Allocation concealment	Blinding of participants and researchers	Blinding of outcome assessment	Incomplete outcome data	Selective reporting	Other bias	Overall risk of bias
Rackow et al. [13]	Unclear	High	High	High	Low	Unclear	Low	High
Metildi et al. [15]	Low	High	High	High	Low	Unclear	Low	High
The SAFE study investigators [6]	Low	Low	Low	Low	Low	Unclear	Low	Unclear
The CRISTAL study investigators [14]	Low	Low	High	Low	Low	Unclear	Low	High
The ALBIOS study investigators [9]	Low	Low	High	High	Low	Low	Low	High

Sequence generation: Rackow et al. [13] used predetermined randomization schedule. Metildi et al. [15] used random numbers. SAFE [6] used a central minimization algorithm. CRISTAL [14] and ALBIOS [9] used computer-generated sequences. Allocation concealment: no detailed allocation concealment was reported by Metildi et al. [15]. Rackow et al. [13] used a predetermined randomization schedule. SAFE [6] used a central minimization algorithm. CRISTAL [14] used sealed envelopes. ALBIOS [9] used a blinded assignment sequence. Blinding: blinding was not reported in Rackow et al. [13], Metildi et al. [15] and the ALBIOS [9] trial. Blinding was performed in the SAFE [6] trial. Data collection members but not researchers were blinded in the CRISTAL [14] trial. Incomplete outcome data: all trials described the follow-up. Selective reporting: the study protocol of ALBIOS [9] was obtained. Other bias: no evidence of other sources of bias.