Global estimates of paediatric tuberculosis incidence in 2013-19: a mathematical modelling analysis

Abstract

Background: Many children who develop tuberculosis are thought to be missed by diagnostic and reporting systems. We aimed to estimate paediatric tuberculosis incidence and underreporting between 2013 and 2019 in countries representing more than 99% of the global tuberculosis burden.

Methods: We developed a mathematical model of paediatric tuberculosis natural history, accounting for key mechanisms and risk factors for infectious exposure (HIV, malnutrition, and BCG non-vaccination), the probability of infection given exposure, and progression to disease among infected individuals. We extracted paediatric population estimates from UN Population Division data, and we used WHO estimates for adult tuberculosis incidence rates. We parameterised this model for 185 countries and calibrated it using data from countries with stronger case detection and reporting systems. Using this model, we estimated trends in paediatric incidence, and the proportion of these cases that are diagnosed and reported (case detection ratio [CDR]) for each country, age group, and year.

Findings: For 2019, we estimated 997 500 (95% credible interval [CrI] 868 700-1 163 100) incident tuberculosis cases among children, with 481 000 cases (398 400-587 400) among those aged 0-4 years and 516 500 cases (442 900-608 000) among those aged 5-14 years. The paediatric CDR was estimated to be lower in children aged 0-4 years (41%, 95% CrI 34-50) than in those aged 5-14 years (63%, 53-75) and varied widely between countries. Estimated CDRs increased substantially over the study period, from 18% (15-20) in 2013 to 53% (45-60) in 2019, with improvements concentrated in the Eastern Mediterranean, South-East Asia, and Western Pacific regions. Over the study period, global incidence was estimated to have declined slowly at an average annual rate of 1·52% (1·42-1·66).

Interpretation: Paediatric tuberculosis causes substantial morbidity and mortality, and these data indicate that cases (and, thus, probably associated mortality) are currently substantially underreported. These findings reinforce the need to ensure prompt diagnosis and care for children developing tuberculosis, strengthen reporting systems, and invest in research to develop more accurate and easy-to-use diagnostics for paediatric tuberculosis in high-burden settings.

Funding: National Institutes of Health.

Figure 1:. Trends in paediatric incidence estimates compared with total case notifications for each WHO region, 2013–19

Solid lines and shaded bands represent modelled point estimates and 95% credible interval for tuberculosis cases in each WHO region. Dashed lines represent reported case notifications.

Figure 2:. Country-level paediatric tuberculosis incidence estimates compared with case notifications in 2019

Countries with no paediatric case notifications in 2019 (25 countries) were excluded from the plot. Of the countries with estimated paediatric incidence higher than 100 cases, eight countries had case detection ratio values lower than 15% (Burkina Faso [7%], Laos [7%], Niger [8%], Guinea-Bissau [10%], Nigeria [11%], Burundi [13%], Mali [13%], and Cameroon [15%]), excluding countries with no reported paediatric case notifications.