“…Most included studies had significant risk of bias (Supplementary Figure 2), 85 with 42 assessed as high risk, 29e31,33e38,40,41,43,45e48,50e56,58,60,62,66e69,71,73e83 11 with some concerns, 32,42,44,49,57,59,63e65,70,72 and 2 with low risk. 39,61 The domain most commonly judged to be high risk was the risk of bias due to deviations from the intended interventions, whereas the domain regarding bias due to missing outcome data had the least risk among included studies.…”