Abstract
BACKGROUND: To compare epidemiological data between recurrent cases after cure (RC), distinguishing relapse from reinfection, after dropout (RD) and new cases (NC) in an ambulatory setting in a TB-endemic country. METHODS: Records of patients who started treatment for pulmonary TB between 2004 and 2010 in a TB clinic were reviewed. Epidemiological data were analyzed. Spoligotyping and MIRU patterns were used to determine relapse or reinfection in 13 RC available. RESULTS: Of the eligible group (1449), 1060 were NC (73.2%), among the recurrent cases, 203 (14%) were RC and 186 (12.8%) were RD. Of RC, 171 (84.2%) occurred later than 6 months after a previous episode, 13 had available DNA, in 4 (30.7%) the disease was attributed to reinfection and in 9 (69.3%), to relapse. Comparing RC to NC, HIV (p < 0.0001) was independent risk factor for RC. When RC and RD were compared, alcohol abuse (p = 0.001) and treatment noncompliance (p = 0.006) were more frequent in RD. CONCLUSIONS: HIV is the sole more important associated factor for RC. This finding points the need to improve the approach to manage TB in order to decrease the chance for exposure especially in vulnerable people with increased risk of developing disease and to improve DOTS strategy to deal with factors associated to treatment noncompliance.