Efforts to control epidemics rely on mathematical and computational models of how infectious agents spread. Such models help to find ways to deter transmission – through vaccination and quarantine, for example, or information campaigns to alter human behaviour. How far a disease spreads reflects a basic competition between the speed of disease spreading, and the countervailing influence of barriers to transmission, including human awareness, which can encourage actions to prevent further spread.
In a new paper, LML Fellow Colm Connaughton and colleagues show that this competition leads to some counter-intuitive effects. They model the interplay of disease spreading and the flow of disease awareness among people, and go beyond previous models in two ways. First, they model dissemination of information using a so-called Maki-Thompson rumour model, which includes the possibility that informed people may fail to prevent further spread because of emotions such as shame. Second, they introduce a parameter to explore the consequences of different relative time scales between disease and rumour propagation. Their results, established in part through extensive numerical simulations, demonstrate that information spreading can reduce how far an epidemic spreads. Self-awareness, which keeps infected individuals aware of their condition, can help reduce disease prevalence.
However, in some circumstances, increasing the speed at which individuals become aware of disease risk can actually help the disease to spread further. If information spreads much faster than the disease, large values of self-awareness can lead unexpectedly to higher disease prevalence. This happens because self-awareness can generate an excessive number of individuals who, despite being aware, don’t use this information to prevent disease transmission. In this way, the work highlights a mechanism which could play an important role in many diseases, such as HIV transmission, where patients are often reluctant to notify their sexual partners.
The paper is available at https://journals.aps.org/pre/abstract/10.1103/PhysRevE.100.032313