The trial is registered with an International Standard Randomised Controlled Trial Number, ISRCTN07601391 (http://www.controlled-trials.com/ISRCTN07601391). These are the results of the 9-year follow up of children re-vaccinated at school age. Baseline data on the individual and cluster characteristics and children excluded from the analysis have been described previously [7]. There were 765 cases of tuberculosis in this analysis: 378 in the intervention group and 387 in the control group, a higher incidence than in previous years given the increase in incidence

of tuberculosis in young adults. Table 1 shows the number of pulmonary and non-pulmonary tuberculosis cases by age of vaccination and by study site. The estimated number of person years of follow up was 1,806,558; 933,107 in the intervention and 873,451 in the control group. The crude incidence of tuberculosis was 41.6 per 100,000 person learn more years in the intervention group and 45.5 per 100,000 person years in the control group (Rate ratio 0.91, 0.79–1.05).

There was no evidence for a design effect when comparing parameters between the naïve and the GEE regression model. Table 1 shows the vaccine efficacy (VE) according to study site and age at diagnosis. Revaccination was protective in Salvador (VE 19%, 3–33%) but not in Manaus (VE 1%, −27 to 23%). In Salvador only children aged <11 years

at vaccination selleck kinase inhibitor where protected (VE 33%, 3–54%). For both cities combined, weak evidence of a protective effect was found (p = 0.08); although the combined measure is of difficult interpretation. Efficacy of BCG revaccination presented a small not significant increase with time of follow up, from 9% (−16 to 29%) at 0–5 years of follow up to 12% (−2 to 24%) at 0–9 years of follow up. Efficacy was almost 20% in Salvador, and practically zero in Manaus; it was higher when given at younger age. Although this finding could be due to chance considering the large and overlapping confidence intervals, it was unexpected: we expected efficacy of revaccination to increase with age at vaccination as efficacy of neonatal BCG decreases. A possible explanation is that infection with Mycobacterium tuberculosis (M. tb) increases with age. In Tolmetin fact, in the study population from Salvador positive PPD results increased from 14.5% in children aged 7–8 years to 28% in children aged 13–14 years [15]. The difference in VE between the two cities was in the direction expected, based on the fact that Manaus is closer to the Equator and presumably has higher prevalence of M. tb and NTMb [3]. Different infection rates with M. tb prior to revaccination could also explain the different vaccine efficacies between the study sites. Infection with M. tb. reduces the protective effect of the BCG vaccine [12].