The spread of infectious diseases through populations depends not just on how contagious a pathogen is, but also on how people respond to it. This paper studies a mathematical model of epidemics where healthy individuals can actively avoid their infected neighbors. The setting is a grid-like network in two or more dimensions, where people occupy the nodes and connections between them can be switched on or off. When a healthy person has an infected neighbor, that connecting link can shut down at some rate, reducing the chance of transmission. The link reopens once the infected neighbor recovers. This captures a simplified version of real-world behavior like social distancing or contact reduction during an outbreak.
The central question is whether an epidemic can survive long-term under this kind of avoidance behavior, or whether avoidance alone is enough to wipe it out. The authors prove that there are two critical thresholds for the transmission rate. If transmission is weak enough, the disease always dies out. If transmission is strong enough, the disease persists indefinitely with some positive probability, even when people are actively avoiding infected contacts. Between these two thresholds, the outcome is uncertain. This shows that avoidance does raise the bar for an epidemic to take hold, but it cannot prevent persistence entirely if the pathogen is sufficiently contagious.
The paper also examines what happens when the number of spatial dimensions grows large, which is a useful mathematical technique for gaining insight into the model's structure. In high dimensions, both critical thresholds shrink in a predictable way, scaling inversely with the number of dimensions. The authors extend their analysis to related epidemic models, including ones where infected individuals enter a temporary isolated or immune state before recovering. The mathematical tools developed here are broadly applicable across this family of models, making the results useful beyond the specific setup studied.