Summary of available clinical treatment and outcome information for multidrug-resistant and extensively drug-resistant tuberculosis (MDR- and XDR-TB) patients
| Characteristics of study participants | MDR-TB patients (with or without resistance to FQN or SLI, not XDR) | XDR-TB1 patients | All patients | ||
| Standardised treatment | Individualised treatment | Total MDR-TB patients | |||
| Total number of studies (cohorts) | 22 (28) | 45 (49) | 67 (77) | 7 (7) | 74 (84) |
| Total number of participants | 5954 | 10 543 | 16 497 | 997 | 17 494 |
| Age median years | 34 | 38 | 38 | 35 | 36 |
| Clinical characteristics2 | |||||
| History of prior treatment with FLD | 1714/1827 | 2859/5030 | 4573/6857 | 50/107 | 4623/6964 |
| History of prior treatment with SLD | 25/613 | 1192/5452 | 1217/6065 | 486/706 | 1703/6771 |
| Patients with cavitation on CXR | 324/508 | 4649/8504 | 4973/9012 | 115/249 | 5088/9261 |
| Patients with HIV co-infection | 1767/5427 | 815/9874 | 2582/15 301 | 529/971 | 3111/16 272 |
| Patients with HIV receiving ART | 483/1767 | 476/815 | 959/2582 | 344/529 | 1303/3111 |
| Positive AFB smears | 1798/2665 | 3995/5674 | 5793/8339 | 264/504 | 6057/8843 |
| Drugs received | |||||
| Pyrazinamide | 5669 | 4948 | 10 617 | 855 | 11 472 |
| Ethambutol | 1908 | 2709 | 4617 | 678 | 5295 |
| Amikacin/kanamycin | 4763 | 4219 | 8982 | 56 | 9038 |
| Capreomycin | 68 | 2352 | 2420 | 823 | 3243 |
| Ofloxacin/ciprofloxacin | 4587 | 5023 | 9610 | 139 | 9749 |
| Moxifloxacin/levofloxacin | 969 | 3686 | 4655 | 150 | 4805 |
| Thiamide | 5927 | 5970 | 11 897 | 731 | 12 628 |
| Cycloserine/terizidone | 2462 | 5044 | 7506 | 720 | 8226 |
| PAS | 214 | 4947 | 5161 | 849 | 6010 |
| 6-month sputum culture conversion | |||||
| Number of studies (cohorts) | 6 (6) | 10 (10) | 16 (16)3 | 4 (4)3 | 19 (19)3 |
| Number of participants | 653 | 2563 | 3216 | 505 | 3721 |
| With conversion | |||||
| Number | 470 | 1515 | 1985 | 99 | 2084 |
| % (95% CI)3,4 | 75 (60–90) | 66 (51–79) | 69 (58–80) | 19 (14–23) | 62 (48–76) |
| EOT outcomes | |||||
| Number of studies (cohorts) | 17 (23) | 33 (38) | 50 (61) | 15 (15)1 | 53 (64)1 |
| Number of participants | 5059 | 9044 | 14 103 | 730 | 14 833 |
| Success5 | |||||
| Number | 2631 | 5797 | 8428 | 193 | 8621 |
| % (95%CI)8 | 52% (50–54) | 64% (63–65) | 60% (58–61) | 26% (23–30) | 58% (57–59) |
| Fail6 | |||||
| Number | 5197 | 992 | 1511 | 2739 | 1784 |
| % (95%CI)8 | 10% (9–11) | 12% (11–13) | 11% (10–12) | 42% (40–44) | 13% (12–14) |
| Fail/relapse6 | |||||
| Number | 519 | 1037 | 1556 | 2739 | 1829 |
| % (95%CI)8 | 10% (9–11) | 13% (12–13) | 12% (11–12) | 42% (40–44) | 13% (12–14) |
| Death6 | |||||
| Number | 839 | 646 | 1485 | 146 | 1631 |
| % (95%CI)8 | 17% (16–18) | 8% (7–9) | 11% (11–12) | 21% (18–25) | 12% (11–13) |
| Lost to follow-up | |||||
| Number | 1045 | 1307 | 2352 | 98 | 2450 |
| % (95%CI)8 | 21% (19–22) | 16% (15–16) | 18% (17–18) | 14% (11–18) | 17% (17–18) |
| All non-success10 | |||||
| Number | 2428 | 3247 | 5675 | 537 | 6212 |
| % (95%CI)8 | 48% (47–50) | 36% (35–37) | 39% (40–41) | 74% (70–77) | 42% (41–43) |
FQN: fluoroquinolone; SLI: second-line injectable; FLD: first-line drug; SLD: second-line drug; CXR: chest radiograph; ART: anti-retroviral treatment; AFB: acid-fast bacilli; PAS: para-aminosalicylic acid; EOT: end of treatment. 1: seven studies (seven cohorts) reported only XDR-TB patients, of which three reported EOT outcomes and three reported 6 month sputum conversion. Another 12 studies reported EOT outcomes stratified by resistance pattern (XDR were separated from MDR). These 12 studies did not stratify the clinical or demographic variables according to resistance pattern, so, for the description of clinical and demographic information of the XDR-TB patients, we used the information from the seven studies that reported only XDR-TB patients. For the pooled EOT analysis, we combined the results from the three cohorts of only XDR-TB patients and the 12 cohorts of XDR-TB patients from the studies that reported stratified results. 2: for clinical characteristics variables we added the denominator, because not all studies reported the clinical variables. The report of these clinical variables was not a criterion for inclusion in the review. 3: one study stratified results of 6-month culture conversion by resistance pattern. Thus, this study is shown in both the XDR column and MDR total column. 4: pooled using Proc Glimmix in SAS – random effects meta-analysis. 5: success was defined as cure or treatment complete. 6: four studies were excluded from the analysis of fail/relapse: two reported fail and death together; one reported fail/death/loss to follow-up together; and the other reported fail/death/loss to follow-up/relapse together. The last two were also excluded from loss to follow-up analysis. One XDR study that reported fail and death together was excluded from the analysis of fail/relapse and death. 7: no studies that reported standardised regimens reported fail separated from relapse. 8: pooled using Proc Glimmix in SAS – fixed effects meta-analysis. 9: no studies that reported XDR reported fail separated from relapse. 10: all non-success includes fail, relapse, death and loss to follow-up.