Background:The timing of highly active antiretroviral therapy (HAART) after a tuberculosis diagnosis in HIV-infected patients can affect clinical outcomes and survival. We compared survival after tuberculosis diagnosis in HIV-infected adults who initiated HAART and tuberculosis therapy simultaneously to those who delayed the start of HAART for at least two months.Methods:The THRio cohort includes 17,983 patients receiving HIV care in 29 public clinics in Rio de Janeiro, Brazil. HAART-naïve patients at the time of a new TB diagnosis between September 2003 and June 2008 were included. Survival was measured in days from diagnosis of TB. We compared survival among patients who initiated HAART within 60 days of TB treatment (simultaneous – ST) to those who started HAART >60 days of TB treatment or never started (deferred – DT). Kaplan–Meier plots and Cox proportional hazards regression analyses were conducted.Results:Of 947 patients diagnosed with TB, 572 (60%) were HAART naïve at the time of TB diagnosis; 135 were excluded because of missing CD4 count results. Among the remaining 437 TB patients, 56 (13%) died during follow-up: 25 (10%) among ST patients and 31 (16%) in DT group (p = 0.08). ST patients had lower median CD4 counts at TB diagnosis than DT patients (106 vs. 278, p < 0.001). Cox proportional hazards utilizing propensity score analysis showed that DT patients were more likely to die (adjusted HR = 1.89; 95% CI: 1.05–3.40; p = 0.03).Conclusion:HAART administered simultaneously with TB therapy was associated with improved survival after TB diagnosis. HAART should be given to patients with HIV-related TB as soon as clinically feasible.

BACKGROUND: Tuberculosis is the most common opportunistic infection among HIV-infected patients in Brazil. Brazil's national policy for HIV care recommends screening for latent tuberculosis (TB) and implementing isoniazid preventive therapy (IPT). OBJECTIVES: We compared physician adherence to TB screening and other prevention and care policies among HIV primary care clinics in Rio de Janeiro City. METHODS: Data on performance of CD4 counts, viral load testing, tuberculin skin testing (TST) and IPT were abstracted from patient charts at 29 HIV clinics in Rio de Janeiro as part of the TB/HIV in Rio (THRio) study. Data on use of pneumocystis jiroveci pneumonia (PCP) prophylaxis were also abstracted from a convenience sample of 150 patient charts at 10 HIV clinics. Comparisons were made between rates of adherence to TB guidelines and other HIV care guidelines. RESULTS: Among the subset of 150 patients with confirmed HIV infection in 2003, 96% had at least one reported CD4 counts result; 93% had at least one viral load result reported; and, PCP prophylaxis was prescribed for 97% of patients with CD4 counts < 200 cells/mm³ or when clinically indicated. In contrast, 67 patients (45%) had a TST performed (all eligible); and only 11% (17) of eligible patients started IPT. Among 12,027 THRio cohort participants between 2003 and 2005, the mean number of CD4 counts and viral load counts was 2.5 and 1.9, respectively, per patient per year. In contrast, 49% of 8,703 eligible patients in THRio had a TST ever performed and only 53% of eligible patients started IPT. CONCLUSION: Physicians are substantially more compliant with HIV monitoring and PCP prophylaxis than with TB prophylaxis guidelines. Efforts to improve TB control in HIV patients are badly needed.