“… 24 , 25 , 62 , 63 , 82 , 83 , 123 , 124 For the two primary outcomes, we included 36 studies (20 on recurrence or relapse 36 , 39 , 43 , 47 , 51 , 64 , 67 , 71 , 84 , 88 , 94 , 96 , 97 , 100 , 109 , 112 , 118 , 123 , 127 , 147 and 16 on mortality during treatment. 32 , 54 , 55 , 57 , 66 , 70 , 74 , 90 , 95 , 99 , 103 , 111 , 120 , 125 , 126 , 135 However, data from eight studies could not be pooled (five on recurrence or relapse 39 , 47 , 67 , 123 , 127 and three on mortality 55 , 66 , 90 ) because the numbers required for computing RRs were not reported. Therefore, 15 studies were included in the TB recurrence or relapse meta-analyses and 13 studies were included in the mortality during treatment meta-analyses ( e-Table 1 ).…”