Scientists examine new way to track outbreaks
Related
Advertisement
New math formula could slow spread of deadly epidemics
5/29/2008 - 5/29/08
Math has long had a reputation for being difficult, but when the next pandemic disease outbreak happens, math could finally redeem itself — and save lives in the process.Scientists at Los Alamos National Laboratories and the Santa Fe Institute say they have found a way to use math to track outbreaks much more quickly than has ever been done in the past.
And the reason that's possible is partly because of the deep flexibility of math, and partly because technologies like the Internet and cell communications have made it much faster to gather data about diseases when they first start to appear, said Luis Bettencourt, a LANL scientist and SFI external professor.
"Using this formula, we could tell very early if something like bird flu jumped to a level where it was rapidly spreading from human to human," Bettencourt said. "And that would let people deal with it much more quickly on the ground."
Most epidemic modeling is done by looking at past outbreaks and using the information to predict how future outbreaks will spread.
It's a solid method, but not very adaptable to a rapidly changing situation, Bettencourt said.
"If you look at an evolving situation, you have rough numbers of cases, but you don't know some things," he said. "With that, you can build a probability model. That's a class of methods that have been used in other cases where it's hard to predict the future, such as weather prediction."
With the mathematical probability model, those dealing with an outbreak can track the number of cases, the methods by which the disease is spreading and other factors in real time.
And each case influences the overarching model, making it more accurate, which in turn makes it easier to react to the disease on the ground. Bettencourt said.
"If you want to know how many hospital beds you need, how much medicine you need, this can help tell you," Bettencourt said.
It could, for instance, project how many cases you might have in two weeks, and how many doctors or how much isolation space you would need to get the outbreak under control, he said.
The formula is also simple enough that doctors can run it on a calculator — or even with a pencil and paper — although the scientists that came up with it are thinking about making a easy-to-use computer program that they can distribute.
It hasn't been used in a real life situation yet, but the scientists tested it retroactively by running it on data from old epidemics, and it worked very well, Bettencourt said.
Bettencourt and co-author Ruy Ribeiro, another Los Alamos scientist, said they hope that agencies like the World Heath Organization and the Centers for Disease Control and Prevention will adopt the idea, although it's still a bit early.
Their paper was recently published on May 14 in the Public Library of Science at plosone.org, and while no health groups have said they plan to use it yet, there has been a lot of interest, Bettencourt said.
If the method is widely adopted, it could help health managers slow progression of disease outbreaks on a global scale, which is important because any disease is really just an airplane trip away, Ribeiro said.
"A program of this type is something that needs to be implemented at a worldwide level to provide an integrated way to respond a priori to an emerging disease threat," Ribeiro said.
Beyond just tracking human cases, the method also can be used to determine how quickly a disease is moving from transmission between various animal species, to transmission from animals to humans and eventually to transmission from human to human, Bettencourt said.
"The most interesting diseases out there are the ones we don't know about yet," Bettencourt said. "Most emerging diseases are diseases of animals. Bird flu, for instance, is mostly a disease of birds. But it has spread into other animals, and it has been transmitted to humans."
Bird flu also, in very rare cases, has been transmitted from human to human, he said.
Those rates are very low right now, but if they were to increase, that could spell trouble, and that's what the math formula could help to quickly determine, Bettencourt said.
"You want to know when a disease jumps from being moderately transmissible between humans to when it becomes substantially transmissible," he said. "If you can see that in real time you can react to it much more effectively."
And because the method can absorb very small amounts of data in real time, it can also tell emergency managers if their strategies toward fighting an outbreak are working, he said.
"Usually public officials try to intervene, but they don't know how well their strategies are working," Bettencourt said. "Generally disease outbreaks are underestimated at the start and then overestimated as they grow. But with this you can basically get a handle on it early."
Considering that bird flu has a 63 percent mortality rate, and that the 1918 pandemic flu outbreak that killed millions had a less than 2 percent mortality rate, every second of early preparation could have a dramatic influence on how many lives are saved, he said.
And that's where math, difficult as it may sometimes be, could very well come to the rescue, he said.
"This could be the start of a whole new world of active epidemiology," Bettencourt said.
Contact Sue Vorenberg at 986-3072 or svorenberg@sfnewmexican.com.
