LAUSR

PURPOSE To determine the short-term and long-term effects of highly active antiretroviral therapy (HAART) on incident tuberculosis (TB) in people living with HIV/AIDS (PLWHA). METHODS From 2000 to 2012, we identified adult PLWHA from Taiwan Centers for Disease Control HIV Surveillance System. All PLWHA were followed up until December 31, 2012, and observed for TB occurrence. Time-dependent Cox proportional hazards models were used to determine the short-term and long-term effects of HAART on incident TB. RESULTS Of 20,072 PLWHA, 628 (3.13%) had incident TB, corresponding to an incident rate of 701/100,000 person-years. After adjusting for potential confounders, PLWHA receiving HAART were more likely to develop TB than those not receiving the drugs (adjusted hazard ratio [AHR] 1.56; 95% confidence interval [CI] 1.18-2.05). While the short-term and long-term effects of HAART on incident TB were considered, HAART was a risk factor for TB development within the first 90 days (AHR 6.06; 95% CI 4.58-8.01) and between 90 and 180 days of treatment (AHR 1.80; 95% CI 1.11-2.94) but was a protective factor after 180 days of HAART use (AHR 0.51; 95% CI 0.39-0.66). CONCLUSIONS HAART is a risk factor for the development of TB in the short term but a protective factor in the long term.