“…Twelve RCTs were judged at overall high risk of bias [17,19,20,24,28,31,33,[35][36][37][38][39], five at overall unclear risk of bias [25,30,32,34,40] and one at overall low risk of bias [23]. None were judged at high risk of bias for random sequence generation or allocation concealment; 13 had high risk of bias for blinding of participants and personnel [17, 19, 20, 23-25, 31-34, 38-40]; eight for blinding of outcome assessment [19,20,24,31,33,[37][38][39]; eight from reporting withdrawals > 10% [17,19,24,28,35,36,39,40]; and five had high risk of bias as some outcomes reported in either the protocol [20,25,28,32] or the methods section [35] were omitted from the results (Supplement 3).…”