“…Three studies [ 32 , 36 , 37 ] were classified as “low risk”, twelve studies [ 16 , 18 , 19 , 28 , 29 , 30 , 31 , 35 , 38 , 39 , 41 , 42 ] as “some concerns”, and four studies [ 17 , 33 , 34 , 40 ] as “high risk” in the overall bias ( Figure 2 ). For bias arising from the randomization process, 12 studies [ 17 , 18 , 19 , 28 , 30 , 31 , 32 , 33 , 36 , 37 , 39 , 42 ] were rated as “low risk”, while 7 studies [ 16 , 29 , 34 , 35 , 38 , 40 , 41 ] were classified as “some concerns.” For bias due to deviations from the intended interventions, all but one study [ 40 ] with “high risk” due to lack of information, and another study [ 38 ], with “some concerns”, were assessed as “low risk.” For the bias caused by missing outcome data, two studies [ 31 , 42 ] were classified as “some concerns”, and four studies [ 17 , 33 , 34 , 40 ] were rated as “high risk” for the following reasons: insufficient information [ 17 , 40 ], and large population of ex...…”