Transmission elasticity in communities hyperendemic for tuberculosis.

BACKGROUND Despite consistently meeting international performance targets for tuberculosis case detection and treatment success, areas where tuberculosis is hyperendemic fail to achieve the predicted epidemiological impact. In this article, we explore the anomalous relationship between defined performance targets and actual reduction in tuberculosis transmission. METHODS In areas where tuberculosis is endemic, poorly ventilated social gathering places such as shebeens (informal alcohol drinking places), minibus taxis, and clinic waiting rooms are all potential transmission hot spots. We modeled the transmission reduction achieved by removal of infectious persons in settings with different tuberculosis prevalence rates to demonstrate the concept of transmission elasticity. We then applied this concept to real-life data from a hyperendemic community in Cape Town, South Africa. RESULTS In a hyperendemic area, reducing the number of infectious people by a given percentage results in a smaller percentage decrease in the annual risk of infection (ARI) compared with a nonendemic area; for example, removing 10% of infectious persons could result in as little as a 5% reduction in the ARI. With use of real-life data and removal of 60% of infectious individuals with tuberculosis, as would be achieved by meeting current performance targets of 70% case detection and 85% cure, the estimated ARI reduction is 50%. CONCLUSIONS The relationship between the number of infectious people removed and the decrease in ARI is nonlinear. The concept of transmission elasticity has important implications for the formulation of universal performance targets, since hyperendemic areas would require more stringent targets to achieve comparable transmission reduction.