Background: Loss to follow up (LTFU) is an obstacle to improving tuberculosis (TB) indicators. The aim of this study was to evaluate the incidence of and risk factors for LTFU from TB treatment by comparison of two prospective cohorts enrolled at the same health centers in southern Brazil: patients who used RH-FDC + Z regimen (2007-2010) and 4-FDC regimen (2018-2019). Methods: Predictors of LTFU were investigated using univariate and multivariate multilevel Cox regression models. Results: Of the 296 cases included, 247 (83.5%) and 49 (16.5%) had cure and LTFU as outcomes, respectively. Patients treated in first period (RH-FDC + Z regimen) and those with substance abuse (drug addicts, alcoholism, and smokers) had a higher risk for LTFU from TB treatment (p < 0.05). Accordingly, the cases treated with RH-FDC + Z regimen showed a HR of 2.25 (95% CI: 1.03 - 4.94, p = 0.042). Conclusions: The main result of this research was the significant reduction in the LTFU from TB treatment after the consolidation of the 4-FDC in Brazil, strengthening the premise that the presentation of drugs in this way improves treatment adherence. Also, substance abuse should be investigated more carefully in TB patients and systematically recorded as risk factors for LTFU, to contribute to better outcomes. Although the TB treatment 4-FDC regimen analyzed is already in use, these new regimens should be implemented with simultaneous pharmacovigilance studies and pragmatic cohort or trial designs to simulate real-world clinical practice, as done by this study.

No abstract