Optimal Duration of Daily Antituberculosis Therapy before Switching to DOTS Intermittent Therapy to Reduce Mortality in HIV Infected Patients: A Duration-Response Analysis Using Restricted Cubic Splines

Compared with thrice-weekly intermittent antituberculosis therapy (ATT), the use of daily ATT during the intensive phase has shown improved survival in HIV infected patients with tuberculosis. However, the optimal duration of daily ATT before initiating intermittent ATT is not well known. In this study, we analysed the mortality of HIV-related tuberculosis according to the duration of daily ATT before switching to thrice-weekly ATT in patients who completed at least two months of treatment in an HIV cohort study. Statistical analysis was performed using Cox proportional hazard models. To relax the linearity assumption in regression models and to allow for a flexible interpretation of the relationship between duration of daily ATT and mortality, continuous variables were modelled using restricted cubic splines. The study included 520 HIV infected patients with tuberculosis and 8,724.3 person-months of follow-up. The multivariable analysis showed that the mortality risk was inversely correlated with the duration of daily ATT before switching to intermittent therapy during the first 30 days of ATT but, after approximately 30 days of treatment, differences were not statistically significant. The results of this study suggest that daily ATT should be given for at least 30 days before switching to intermittent ATT in HIV infected patients with tuberculosis.