| Methods |
1583 individuals "randomized," stratified by clinic; blinding: providers no, participants no, assessors yes. 115 (7%) lost to follow‐up; intention to treat analysis. |
| Participants |
Patients attending HIV/AIDS clinics in US, Mexico, Haiti and Brazil. Inclusion criteria: HIV positives, age >= 13 years, >=5mm induration to 5 U PPD or documented history of positive test, informed consent. Exclusion Criteria: Active TB, current treatment with fluoroquinolones or history of >2 mo treatment with anti‐TB drugs, intolerance to study drugs, acute hepatitis or peripheral neuropathy, pregnancy. |
| Interventions |
1) INH 300mg plus pyridoxine 50mg daily for 12 months.
2) RIF 600mg plus PZA 20mg/kg, daily for 2 months. |
| Outcomes |
1) Active TB: positive culture from any source. 2) Probable TB: clinical evidence from a physical examination. 3) Clinical progression of HIV disease: first occurrence of AIDS defining condition. 4) Death. 5) Adverse Effects. ‐‐ Mean duration of follow‐up 37months. |
| Notes |
Use of ART at baseline 35.8% in INH arm and 36.8% in RIF/PZA arm. "Progression of HIV was not reliably diagnosed in Haiti." |
| Risk of bias |
| Bias |
Authors' judgement |
Support for judgement |
| Allocation concealment? |
Unclear risk |
B ‐ Unclear |
| Blinding?
All outcomes |
Low risk |
Blinding of assessors only, no blinding of providers or participants |
| Incomplete outcome data addressed?
All outcomes |
Low risk |
7% lost to follow‐up |
