Emergent heterogeneity in declining tuberculosis epidemics

Abstract

Tuberculosis is a disease of global importance: over 2 million deaths are attributed to this infectious disease each year. Even in areas where tuberculosis is in decline, there are sporadic outbreaks which are often attributed either to increased host susceptibility or increased strain transmissibility and virulence. Using two mathematical models, we explore the role of the contact structure of the population, and find that in declining epidemics, localized outbreaks may occur as a result of contact heterogeneity even in the absence of host or strain variability. We discuss the implications of this finding for tuberculosis control in low incidence settings.

Fig. 1

The spatial and delay DE models with observed data for incidence after 1950. Dashed lines are the spatial model and solid lines are the delay model. (a) Incidence (cases in 100,000): 1900−2000. (b) Incidence (cases in 100,000): post-treatment era.

Fig. 2

Eigenvalues showing stability switch. (a) Disease-free. (b) Endemic.

Fig. 3

The ratio of exogenous reinfection to primary activation for the network model with D = 1 and the differential equation model for several values of the transmission parameter, β.

Fig. 4

The variance of local latency compared to the predicted variance without spatial clustering.

Fig. 5

The time evolution of local latency for D = 1 and 10. Height indicates the number of individuals with the given number of latent neighbors at the corresponding time. At 50 years we decrease the transmission parameter to mimic the decline in TB between 1900 and 1950, and at 100 years (1950), we introduce antibiotic treatment. (a) D = 1. (b) D = 10.