Baseline characteristics of included participants

Years of enrolment2005–20112008–20112008–20102013–20152014–2016
Participants included in meta-analyses493150656424
Age years33.6±12.935.131 (27–38)33.7±14.435.2±14.4
Female150/493 (30.4)73/150 (48.7)12/65 (18.5)35/64 (54.7)13/24 (54.2)
Primary MDR4/493 (0.8)1/150 (0.1)1/65 (1.5)46/60 (76.7)20/24 (83.3)
HIV-positiveNR30/150 (20.0)1/58 (1.7)0/4416/24 (66.7)
Smear-positive475/493 (96.3)150/150 (100)54/65 (83.1)28/61 (45.9)12/23 (52.2)
Chest radiograph cavities99/493 (20.1)NR23/65 (35.4)26/60 (43.3)NR
Baseline resistance
 Fluoroquinolones#65/485 (13.4)0/1331/65 (1.5)0/501/15 (6.7)
 Kanamycin0/4850/1350/6513/49 (26.5)0/14
 CapreomycinNRNRNR2/48 (4.2)NR
 Prothionamide/ethionamide82/466 (17.6)NR7/65 (10.8)NR5/14 (35.7)
 Pyrazinamide99/240 (41.3)NRNR32/38 (84.2)10/14 (71.4)
 Ethambutol321/493 (65.1)NR45/65 (69.2)30/43 (69.8)12/17 (70.6)

Data are presented as n, mean±sd, group mean, median (interquartile range) or n/n (%), unless otherwise stated. Denominators less than the total participants are due to missing data. MDR: multidrug-resistant; NR: not reported; DST: drug susceptibility test; MIC: minimum inhibitory concentration; TB: tuberculosis. #: in all studies, DSTs for ofloxacin were performed to identify resistance to fluoroquinolones. In Bangladesh, 163 participants also had DSTs for gatifloxacin performed. Gatifloxacin resistance was defined as low level if MIC 0.5–1 mg·L−1 and high level if MIC ≥2 mg·L−1 [7]. Of the 65 participants with fluoroquinolone-resistant TB from Bangladesh, the breakdown of the resistance profile was as follows: n=13 ofloxacin-susceptible with low-level gatifloxacin resistance, n=20 ofloxacin-resistant with low-level gatifloxacin resistance, n=27 ofloxacin-resistant with high-level gatifloxacin resistance, and n=5 ofloxacin-resistant and gatifloxacin DSTs not performed. We classified five participants from Bangladesh with ofloxacin-resistant and gatifloxacin-susceptible MDR-TB as fluoroquinolone-susceptible.