The widely used criminology technique, called geographic profiling, helps investigators narrow a search by pinpointing high-priority targets among thousands of potential locations. In an upcoming International Journal of Health Geographics, researchers demonstrated the technique’s usefulness by identifying the sources of a recent malaria outbreak in Cairo and reconstructing an infamous cholera outbreak in Victorian London. Applying the technique to infectious diseases could help focus interventions, perhaps preventing the spread of disease while saving time and money.
“I think this has a lot of promise,” says disease ecologist Richard Ostfeld of the Cary Institute of Ecosystem Studies in Millbrook, New York. “It’s a very interesting application of a criminological tool to epidemiology.”
When hunting criminals, geographic profiling uses the sites of connected crimes to figure out where a criminal might live. Pioneered by criminologist Kim Rossmo, a former Vancouver police officer now at Texas State University-San Marcos, the method is based on a criminal’s tendency to take a Goldilocks-like approach when selecting where to commit a crime — a location that’s not too close to home, not too far, but just right.
Rossmo, a coauthor of the new study, developed an algorithm that incorporates this notion in two parts. The crime is less likely to be committed in the criminal’s buffer zone — the immediate vicinity of his or her home or work — because detection is riskier and opportunities may be few. And the likelihood of a crime site decays with distance, because travel requires time, effort and money.
“I’m based in London,” says study coauthor Steven Le Comber of Queen Mary, University of London. “So I’m not going to pop up to Inverness [in the far reaches of Scotland] to murder someone. But, equally, I don’t want to commit crimes on my own doorstep.”
The math behind geographic profiling also incorporates the idea that all distances are not created equal — highways are easier to traverse than a congested downtown. All these measures then generate a map of places the offender is likely to live, which is overlaid on a map of a search area. Unlike geospatial techniques that designate a central point from which a search radiates equally outward, geographic profiling pinpoints highly probable locations, even if they are at opposite ends of the search area.
Le Comber and his colleagues applied geographic profiling to a recent malaria outbreak in Cairo. Of 59 water bodies where mosquito larvae were found, only eight contained those species that are the most dangerous carriers of the disease. Knowing only the locations of the outbreak’s 139 malaria cases, geographic profiling correctly put six of these eight sites in the most infectious 2 percent of the 59.
“I have to say, it impressed us,” says Le Comber. “We thought it would work, but it was a bit stunning.”
The team also used the technique on the 1854 London cholera outbreak, from which doctor John Snow famously created a map of cholera deaths. This led Snow to the Broad Street pump as a source of the disease, and launched the modern field of epidemiology. Based on 321 deaths, geographic profiling also ranked the Broad Street pump as the most likely origin of the outbreak, the researchers report.
In some instances, such as diseases that have secondary outbreaks, geographic profiling might not work as well, says Ostfeld. But it could help yield more accurate maps for ecological questions, he says, such as how species ranges might respond to climate change. Researchers are already using the method to study how sharks hunt and the movement of bats and bees.
Read more at Wired Science
No comments:
Post a Comment