Gillespie 2014.

Methods	Study design: multi‐centre, randomized, parallel‐group, double‐blind (participant, care provider, investigator, outcomes assessor), 3‐armed, placebo‐controlled, non‐inferiority trial Study period: January 2008 to February 2014 Recruitment sites: 47 sites in 9 countries Countries where the trial was undertaken: South Africa, India, Tanzania, Kenya, Thailand, Malaysia, Zambia, China, Mexico Length of follow‐up: 18 months after randomization (1 year after treatment completion)
Participants	No. of participants randomized: 1931 Interventions: 1291 (636 to ethambutol group; 655 to isoniazid group) Control: 640 . Age: > 35 years, 37% in isoniazid group, 39% in ethambutol group, and 40% in control group Gender: male 70% in ethambutol group, 68% in isoniazid group, 70% in control group Inclusion criteria: 2 sputum specimens positive for tubercle bacilli on direct smear microscopy, of which 1 was confirmed by the REMoxTB study laboratory at the local laboratory No history of previous anti‐tuberculosis chemotherapy Aged 18 years and older Firm home address that is readily accessible for visiting and willingness to inform the study team of any change in address and follow‐up period Agreement to participate in the study and to give a sample of blood for HIV testing Negative pregnancy test (women of childbearing potential) Pre‐menopausal women must be using a barrier form of contraception or must be surgically sterilized or have an IUCD in place Laboratory parameters performed at least 14 days before enrolment Serum aspartate transaminase (AST) and alanine transaminase (ALT) activity less than 3 times upper limit of normal Serum total bilirubin level less than 2.5 times upper limit of normal Creatinine clearance (CrCl) level greater than 30 mL/min Haemoglobin level at least 7.0 g/dL Platelet count at least 50 x 10⁹ cells/L Serum potassium greater than 3.5 mmol/L Exclusion criteria: Patients unable to take oral medication Previously enrolled in this study Receiving any investigational drug in the past 3 months or an antibiotic active against M tuberculosis Pregnancy or breastfeeding Any condition that may prove fatal during the first 2 months of the study period Severe tuberculosis with high risk of a poor outcome (e.g. meningitis) Pre‐existing condition likely to prejudice the response to, or assessment of, treatment; a condition likely to lead to uncooperative behaviour Contraindication to any medications in the study regimens Congenital or sporadic cardiac syndrome or taking medications that could result in QTc prolongation Patients already receiving antiretroviral therapy Weight less than 35 kg HIV infection with CD4 count less than 250 cells/μL End‐stage liver failure (class Child‐Pugh C) Patients whose initial isolate was shown to be multiple drug resistant or monoresistant to rifampicin, or to any fluoroquinolone Proportion with HIV seropositivity: 7% overall (and in intervention and control groups) Proportion with cavitation: 71% overall (69% and 70% in intervention groups and 72% in control groups) Baseline drug resistance: isoniazid: 7% overall (6% in control arm and 7% in each intervention arm); pyrazinamide: 2% overall (1% in each intervention arm)
Interventions	Interventions: 4‐month (17‐week) ATT regimen Isoniazid group (moxifloxacin for 17 weeks substituting ethambutol): N = 655; 568 eligible, 514 completed (78% of those randomized; 91% of those eligible) 8 weeks of moxifloxacin, isoniazid, rifampicin, pyrazinamide, + ethambutol placebo administered daily, followed by 9 weeks of moxifloxacin, isoniazid, and rifampicin, followed by 9 weeks of isoniazid and rifampicin placebo Ethambutol group (moxifloxacin for 17 weeks substituting isoniazid): N = 636; 551 eligible, 524 completed (82% of those randomized; 91% of those eligible) 8 weeks of moxifloxacin, ethambutol, rifampicin, pyrazinamide + isoniazid placebo administered daily, followed by 9 weeks of moxifloxacin and rifampicin + isoniazid placebo daily, followed by 9 weeks of isoniazid and rifampicin placebo Control: 6‐month (26‐week) ATT regimen: N = 640; 555 eligible, 510 completed (80% of those randomized; 92% of those eligible) 8 weeks of isoniazid, rifampicin, ethambutol, pyrazinamide, and moxifloxacin placebo given daily, followed by 9 weeks of isoniazid, rifampicin, and moxifloxacin placebo given daily, followed by 9 weeks of isoniazid and rifampicin Dosage: Moxifloxacin 400 mg, isoniazid 300 mg; rifampicin, ethambutol, and pyrazinamide were dosed based on weight
Outcomes	Outcomes reported and used in this review: Relapse within 18 months after randomization (relapse strains were those shown to be identical on 24‐locus MIRU analysis) Death from any cause Rates of treatment discontinuation Sputum smear/culture positivity at 8 weeks Proportion with bacteriologically or clinically defined failure Serious adverse events Other adverse events Outcomes sought but not reported: Development of secondary drug resistance to anti‐tuberculosis drugs, identified by drug susceptibility testing Outcomes reported but not used in this review: Composite unfavourable outcome (clinical or bacteriologic failure or relapse within 18 months after randomization) (non‐inferiority was defined as a between‐group difference of less than 6 percentage points in the upper boundary of the 2‐sided 97.5% Wald confidence interval for the difference in proportion of patients with an unfavourable outcome) Re‐treated for tuberculosis Time to first culture‐negative sputum Pharmakokinetic data (substudy reported separately)
Notes	Funding: Global Alliance for TB Drug Development (supported by multiple international donor agencies and local agencies and institutions in participating countries). Bayer Healthcare donated moxifloxacin and Sanofi donated rifampicin Treatment supervision: treatment was given daily and was observed according to guidelines at the study site Follow‐up method: following screening and baseline visits, there were 8 weekly visits followed by 8 visits until 18 months after randomization. Safety analysis was performed at the screening visit and thereafter at weeks 2, 8, 12, and 17 Trial registration ID: NCT00864383 (retrospectively registered: registered March 2009; study start January 2008) Acronym: REMoxTB Comment: data from both moxifloxacin‐containing shorter regimens were combined and compared with data from the standard treatment regimen
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote from report: "Randomization was performed with the use of lists with blocks of variable sizes that were stratified according to the patient weight group and study centre"
Allocation concealment (selection bias)	Low risk	Quote from report: "During randomization, patients were assigned a unique study number selected sequentially from the appropriate randomization list that corresponded to the treatment pack allocated" Quote from report: "Only statisticians who were responsible for preparing the reports for the independent data and safety monitoring committee and essential manufacturing and distribution staff members had access to the list of identifiers matched to the intervention"
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote from study protocol: "This will be a blinded study with matching placebo for each of the study medicines except for pyrazinamide" Quote from trial registration document: "Masking: quadruple (participant, care provider, investigator, outcomes assessor)"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote from trial registration document: "Masking: quadruple (participant, care provider, investigator, outcomes assessor)"
Incomplete outcome data (attrition bias): At the end of ATT (Treatment failure, positive sputum culture, treatment discontinuation, adverse events)	Low risk	Quote from report: "Of the 1931 patients who underwent randomization, 89% in the isoniazid group, 92% in the ethambutol group, and 89% in the control group met the requirements for treatment adherence, which was based on receipt of approximately 80% of the assigned regimen" Comment: the modified‐intention‐to‐treat analysis used included 87% of those randomized to the combined moxifloxacin‐containing ATT regimens and 87% of those randomized to standard regimens. A sensitivity analysis included 94% of those randomized to both regimens
Incomplete outcome data (attrition bias): At the end of follow‐up (Relapse, deaths)	Low risk	Of those randomized, 91% of those allotted to the 2 moxifloxacin combination therapy arms and 92% allotted to control treatment were included in the modified intention‐to‐treat analyses. Results of the per‐protocol and modified intention‐to‐treat analyses were consistent
Selective reporting (reporting bias)	Low risk	Although the trial was retrospectively registered, all pre‐stated outcomes listed in the trial registration document and protocol were published with no evidence of selective reporting
Other bias	Low risk	Quote from report: "Bayer Healthcare donated moxifloxacin, and Sanofi donated rifampin. Neither company had any role in the study design, data accrual, data analysis, or manuscript preparation. Representatives of Bayer Healthcare reviewed the manuscript but did not suggest revisions"