An agent-based modelling approach to investigate the impact of gender on tuberculosis transmission in Uganda

TL;DR

This study develops an age- and gender-stratified agent-based model, EpiabmTB, adapted from CovidSim/Epiabm, to quantify how gender affects TB transmission in Kampala, Uganda. Through calibrated simulations and counterfactual scenarios, it demonstrates that both within-host factors (e.g., cavitary TB and progression from latent to active TB) and between-host factors (notably gendered contact patterns and diagnosis delays) contribute to a higher TB burden in males. The analysis of super-spreaders reveals a heavy-tailed transmission distribution, with a minority of infectors driving the majority of cases, and that these infectors are more likely to be male and cavitated. Four counterfactuals consistently reduce the male-to-female TB case ratio, with equalizing within-host progression producing the largest reduction in total cases, highlighting actionable targets for TB control, such as rapid diagnosis and addressing cavitation risk. Overall, the work provides a data-driven framework to inform Uganda's TB control policy and city-level interventions in Kampala.

Abstract

Tuberculosis (TB) is an airborne disease caused by the pathogen Mycobacterium tuberculosis. In 2023, it returned to being the leading cause of death from an infectious agent globally, replacing COVID-19; in the nineteenth century, one in seven of all humans died of tuberculosis. More than 10 million people are diagnosed with TB every year. The majority of cases in adults occur in males (62.5% of all global adult cases in 2023, compared to 37.5% in females). The main reasons for males suffering from a higher burden of global TB cases, compared to females, may be in large part due to population-scale factors, such as employment type, the quantity and type of social contacts they make, and their health-seeking behaviours (e.g. differences in diagnostic and treatment delays between genders). To investigate which population-scale factors are most important in determining this higher TB burden in males, we have developed an age- and gender-stratified, spatially heterogeneous epidemiological agent-based model. We have focused specifically on Kampala, the capital of Uganda, which is a high-burden TB country. We considered counterfactual scenarios to elucidate the impact of gender on the epidemiology of TB. Setting disease progression parameters equal between the genders leads to a reduction in both male-to-female case ratio and total case numbers.

Figures (24)

• Figure 1: Infectious disease states used in our TB epidemiological model and the potential transitions between them. The arrow into the Healthy compartment from the left of the figure represents births: newborns always enter the Healthy compartment at birth. See text for full description.
• Figure 2: Tuberculosis diagnosis delay distributions for females and males from Kampala, Uganda. The distributions are kernel density estimates with standard normal kernel functions and positive support using data from Nsawotebba2025Nsawotebba2025
• Figure 3: Log-normal distribution of Ugandan business sizes (by number of employees), with mean and standard deviation taken from Goyette2014.
• Figure 4: Assumed distribution of Ugandan household sizes, based on UBOS2024 and Jarosz2021. Households of size 1 to 5 have probabilities taken from UBOS2024; households of size 6 or more are calculated based on the method given in Jarosz2021.
• Figure 5: Summary plots of the male-to-female TB case ratio and total TB case numbers for all 300 simulations of the EpiabmTB model, as well as the TB cases stratified by gender. The shaded regions and the dashed lines at their boundaries indicate the 95% confidence intervals. The solid lines contained within these regions indicate the mean amounts of the random variable in question.