MANCHESTER, U.K.--The starts and stalls of a forest fire as it jumps across a dry plain or balks at wet grass have helped scientists forecast the spread of raccoon rabies across New York state. The model is among the first to successfully gauge both how fast a disease will move and what route it is most likely to take.
Much of the research into predicting the spread of disease has focused on the pace of infection. Other models forecasted the progress from place to place, by presuming a steady diffusion of cases from the initial outbreak. But few models have taken into account the actual landscape. In the case of rabies, relevant features might include a river that blocks infected animals or a garbage truck that shuttles rabid raccoons from town to town.
Forest fire models rely heavily on landscape, so population biologist Leslie Real of Emory University in Atlanta and colleagues tried adapting them for disease. They plugged in data from the Centers for Disease Control and Prevention on a variant of rabies in raccoons that was first seen in Virginia and West Virginia in 1977 and has since moved steadily north. Real and collaborators fine-tuned the model with data from the disease's spread through Connecticut. They also factored in barriers such as rivers, as well as valleys, which they reasoned could funnel animals in a particular direction. Applying the model retrospectively, the researchers accurately predicted how rabies spread through New York state, Real's colleague ecologist Colin Russell told the meeting of the British Ecological Society here on 11 September.
“The thing I find most interesting is the model's ability to detect how geographic features of the landscape ... either block or focus the movement pattern of the disease,” says Andrew Dobson, an infectious disease ecologist at Princeton University in New Jersey. Dobson adds that one next step is to use the model to predict ways to stop the spread of the disease, perhaps by setting up barriers at key points. Meanwhile, Real and others have begun modifying the model for other diseases that are spread through close contact, such as HIV and tuberculosis.