“…Only 15 trials reported random numbers generated by computer and random enumeration of closed envelopes to generate random sequences [ 32 , [34] , [35] , [36] , [37] , 40 , 42 , 44 , 50 ]. 16 trials described the low risk of allocation concealment [ 32 , [34] , [35] , [36] , [37] , [38] , [39] , [40] , [41] , [42] , 44 , 47 , 49 , 50 , 52 ]. Three trials [ 33 , 34 , 48 ] had a high risk of bias when it came to incomplete outcome assessment.…”