Forecasting the path of a raccoon rabies epidemic.

In the mid-1970s, a new strain of raccoon rabies started spreading throughout the eastern United States. Raccoons caught in Florida had been released along the West Virginia--Virginia border to replenish hunting stocks, and some of the imports carried a rabies variant that caused an outbreak in local populations. The disease has been steadily spreading among raccoons ever since. In 1990, raccoons topped the list of most often reported rabid mammal.

Controlling this re-emerging public health threat depends on predicting the spatial dynamics of the disease--where new outbreaks might occur and how the virus might spread. Toward this end, a group of researchers have been working on probabilistic simulation models that calculate the effects of various factors such as local transmission rates between townships, ecological barriers to transmission, and long-distance "translocation" rates between townships. (The deliberately released Florida raccoons were one such translocation, but raccoons have also been known to hitch rides on garbage trucks.) These models previously accurately reflected rabies spread in both Connecticut and New York. In a new study reported in PLoS Biology, Leslie Real, of the Department of Biology and Center for Disease Ecology at Emory University, and co-authors apply their model to the likely spread of rabies in Ohio--a potential gateway for spread throughout the Midwest. They find that raccoon rabies could spread throughout the state in just three years.

One strategy for limiting rabies spread is to establish vaccine corridors by distributing vaccine baits--vaccine doses hidden in fishmeal--to wild raccoons. This cordon sanitaire strategy limited rabies in Ohio to sporadic cases from 1997 until 2004, when a rabid animal was detected in northeastern Ohio--11 kilometers beyond the buffer zone. The authors had previously shown that local transmission was significantly reduced when townships were separated by geographical barriers--the Connecticut River in Connecticut and the Adirondack Mountains in New York. In modeling the likely transmission path in Ohio, the authors incorporated the likely effect of Ohio's five major rivers on transmission from local points along the Pennsylvania or West Virginia border.

Given Ohio's topography (three of its rivers run along the southern and eastern border) and a single point of emergence in the northeast, the authors adjusted their simulations to estimate the potential impact of translocations. Even without the occasional garbage truck ride, because of the lack of ecological barriers in central Ohio, the simulations predict that rabies will spread far faster in Ohio than in New York and Connecticut.

Factoring in those garbage truck rides, the scenario is considerably bleaker: rabies would take just 33 months to spread across central Ohio--compared with 48 months to cross the much smaller state of Connecticut--and cover the state in 41 months. This transmission rate--100 kilometers per year--significantly surpasses previous estimates, which range from 30 to 60 kilometers per year. The potential for such rapid spread, if unchecked, "is quite alarming," the authors warn. But they also point out that the path of a real epidemic would likely fall somewhere between these two scenarios, given the unpredictable nature of translocations.

The authors' simulations provide a valuable resource for anticipating alternate scenarios and preparing multiple game plans to prevent or contain outbreaks. They also indicate the best sites for establishing a new vaccine barrier. Given how fast raccoon rabies could spread. Real and colleagues make a strong case that halting its westward march depends on a strategy based on early detection and high-powered intervention programs.

[Adapted from "Forecasting the Path of a Raccoon Rabies Epidemic," PLoS Biol 3(3): e115, www.plosbiology.org. (2005).]