Montejo 1993a.

Methods	Randomized controlled trial Duration: 1980‐83
Participants	Number: about 200 patients enrolled Inclusion criteria: ≥ 14 years of age, a clinical picture consistent with a diagnosis of brucellosis, isolation of germs from clinical specimens, and/ or titers of the standard tube agglutination test (STAT) of 1/160. Exclusion criteria: antecedents of brucellosis in the previous year, the presence of serious associated illness, pregnancy, a reported allergy to one or more of the antimicrobial agents used in this study, and a diagnosis of endocarditis, spondylitis, or CNS affected by Brucella.
Interventions	1‐1200 mg of rifampicin once a day for 7 days and then 600 mg/d for 21 days together with 200 mg of doxycycline/d for 28 days 2‐160 mg of trimethoprim plus 800 mg of sulfa‐methoxazole every 8 hours for 10 days, then the same dose every 12 hours until the end of the 4‐week period, and finally 80 mg of trimethoprim and 400 mg of sulphamethoxazole every 12 hours until the end of the 6‐month treatment period 3‐200 mg of doxycycline/d for 6 weeks
Outcomes	1‐relapse 2‐therapeutic failure 3‐drug regimen change due to non‐compliance
Notes	Location: Hospital de Cruces, Baracaldo, Vizcaya, Spain Setting: not described Source of funding: not mentioned
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Use of a table of random numbers
Allocation concealment (selection bias)	Low risk	The results of this random distribution were placed in sealed, numbered envelopes, which were opened once it was established that the patient fulfilled the criteria for inclusion in the study.
Blinding (performance bias and detection bias) All outcomes	High risk	Open study
Incomplete outcome data (attrition bias) All outcomes	High risk	In total, 5 patients were excluded due to the allergy to medications, 5 because of receiving treatment in the previous year, 4 due to brucellosis complications, and 13 due to loss to follow‐up. At one‐year follow‐up, 65 patients were analysed to the regimen 1, 64 to the regimen 2, and 71 to the regimen 3. The proportion of missing outcomes compared with observed event was considerable.
Selective reporting (reporting bias)	Low risk	Important outcomes are reported
Other bias	Low risk