HKCS/BMRC 1992

Methods	Randomized controlled trial; matched pairs design Generation of allocation sequence: unclear Allocation concealment: central randomization Blinding: assessor only Inclusion of all randomized participants: 64.7% (11/17) in both arms
Participants	Number: 34 enrolled Inclusion criteria: tuberculosis patients failing first‐line regimen and resistant to rifampicin, isoniazid, and streptomycin on susceptibility testing 34/88 screened were eligible: 18 were susceptible to 1 or more of above drugs; 13 culture negative; 13 died before entry; 6 declined to participate; and 4 were on ofloxacin
Interventions	1. Rifabutin: 450 mg if < 50 kg and 600 mg if > 50 kg daily for 12 months 2. Rifampicin 450 mg if < 50 kg and 600 mg if > 50 kg daily for 12 months Companion drugs: 1 to 3 other drugs given daily chosen from: pyrazinamide (1.5/2.0 g); ethambutol (25 mg/kg for 2 months followed by 15 mg/kg thereafter); ethionamide/prothionamide (500 mg); kanamycin (750 mg); capreomycin (750 mg); and para‐aminosalicylic acid (10 g). Those who failed rifampicin could be switched to rifabutin, while those failing rifabutin could be switched to ofloxacin (800 mg daily)
Outcomes	1. Cure/failure at 12 months 2. "Temporary response" (ie culture conversion of limited duration) Not included in this review: 3. Smear and culture status at 2 months
Notes	Location: Hong Kong Supervision: included a mixture of inpatients and outpatients; duration of hospitalization not stated Follow up: sputum collected every 2 weeks for first 2 months then monthly; follow up was 65% at 12 months; 3 participants were culture negative at baseline, and the rest were excluded due to protocol violations Other: "temporary" responses were observed in 18/22 and did not differ between the regimens No power calculation was presented for any outcome measure