Background

Few studies have evaluated tuberculosis control in children and adolescents. We used routine tuberculosis surveillance data to quantify age- and HIV-stratified trends over time and investigate the relationship between tuberculosis, HIV, age and sex.
Methods

All children and adolescents (0-19 years) routinely treated for drug-susceptible tuberculosis in South Africa and recorded in a de-duplicated national electronic tuberculosis treatment register (2004-2016) were included. Age- and HIV-stratified tuberculosis case notification rates (CNRs) were calculated in four age bands: 0-4, 5-9, 10-14 and 15-19 years. The association between HIV infection, age and sex in children and adolescents with TB was evaluated using multivariable logistic regression.
Results

Of 719,400 children and adolescents included, 339,112 (47%) were 0-4-year-olds. The overall tuberculosis CNR for 0-19-year-olds declined by 54% between 2009 and 2016 (incidence rate ratio [IRR]=0.46, 95% confidence interval [CI] 0.45-0.47). Trends varied by age and HIV, with the smallest reductions (2013-2016) in HIV-positive 0-4-year-olds (IRR=0.90, 95%CI 0.85-0.95) and both HIV-positive (IRR=0.84, 95%CI 0.80-0.88) and HIV-negative (IRR=0.89, 95%CI 0.86-0.92) 15-19-year-olds. Compared to 0-4-year-old males, odds of HIV co-infection among 15-19-year-olds were nearly twice as high in females (adjusted odd’s ratio [aOR]=2.49, 95%CI 2.38-2.60) than in males (aOR=1.35, 95%CI 1.29-1.42).
Conclusions

South Africa’s national response to the HIV epidemic has made a substantial contribution to the observed declining trends in tuberculosis CNRs in children and adolescents. The slow decline of tuberculosis CNRs in adolescents and young HIV-positive children is concerning. Understanding how tuberculosis affects children and adolescents beyond conventional age bands and by sex, can inform targeted tuberculosis control strategies.