Matt Keeling and Chris Gilligan turned the study of how plague spreads in humans on its head: by analysing rat numbers. Rodent populations are important because fleas usually spread the plague bacteria
Such research may sound obvious, but this is the first time that rat populations have been carefully modelled to understand bubonic plague in people. "This is one of those pieces of work that you say 'oh why didn't I think of that' after it is published," comments Graham Medley, an infectious disease epidemiologist at Warwick University, UK.
The computer model shows that plague can persist in very low rat populations (about 60,000) and spreads to humans only when some random event reduces rodent numbers. And it simulates what happens when there is a drastic reduction in rat numbers, caused by anything from mass poisonings to a strangely dressed man playing a hypnotic pipe.
Broadly speaking, such a cull will not succeed if the plague has already spread to the human population. If fleas are biting people then it may be better to let the rats run free, giving the insects more rodent targets. Keeping the rat population permanently low seems the only way to avoid serious outbreaks. Vaccination is generally ineffective, Keeling warns, because direct person-to-person transmission is rare.
Although bubonic plague has been eradicated in many parts of the world, outbreaks still kill about 4,000 people each year and antibiotic-resistant strains have emerged. Historically, plague was a massive killer - wiping out one-third of Europe's population when it arrived in the fourteenth century for instance. But many have puzzled over how the disease persisted despite long disease-free periods and strict quarantine.
Keeling says that the new model explains this enigma: the disease can survive in very small rat populations. And quarantine is ineffective because the bugs can be present in a town's rats but not show up in its people - so future outbreaks are not caused by an influx of fresh disease-carrying rats. The model results are announced in
"The study is extremely interesting because it may indicate a more general truth that could be applied to other vector-borne diseases," Medley says. Although a shortage of accurate historical data makes it difficult to verify the model's conclusions, he adds.
Keeling agrees: "Records often just say that someone died of a fever, which was attributed to plague." Still, the model does show close agreement with historical mortality data for London.
Looking ahead, he says that the model has important implications for public health. Bubonic plague is found in wild rodents in many parts of the world, he says, and careful monitoring of urban rat populations could make future outbreaks less likely.