Tuesday, April 23, 2013

Human mobility data may help curb urban epidemics

 
Emory post-doc Donal Bisanzio created a data visualization, above, of the movement routines of people in Iquitos, Peru.

By Carol Clark

Residents of cities like New York and London tend to move about in fairly predictable routines, following the same routes between their jobs and schools each day. When it comes to a city in the developing world, however, human movement is much more varied, a finding with important implications for controlling an infectious disease pandemic.

The Public Library of Science (PL0S One) published the first major analysis of daily human mobility in a resource-poor city, led by scientists at Emory University’s Department of Environmental Studies.

The researchers used GPS technology to quantify the movement and contact dynamics of nearly 600 residents of Iquitos, Peru. They applied the data to create a computer simulation for predicting the transmission rate of a flu virus.

“We found that the irregular movement of people in Iquitos increases the probability of flu transmission by 20 percent, compared to cities in developed nations,” says lead author Gonzalo Vazquez-Prokopec, an Emory disease ecologist.

The study authors are making their data estimates and simulation methods publicly available, so that other researchers can conduct further experiments and build on their work.

“It is estimated that more than 90 percent of the mortality from a potential influenza pandemic would occur in developing countries, where vaccine and antiviral stockpiles are minimal,” the study authors write. “The lack of detailed models to estimate infectious disease transmission dynamics in such settings limits the ability to enforce containment measures or plan emergency preparedness strategies.”

Rather than commuting to a single workplace, poorer residents of Iquitos often work several jobs, such as driving a three-wheeled mototaxi, or selling produce at multiple markets. Photo via Wikipedia Commons.

Most previous data on human mobility, drawn from cities in North America and Europe, shows that urbanites visit an average of two to four locations daily.

In Iquitos, human movement is much more fluid.

Full-time jobs are scarce for the population of 500,000 living along the left bank of the Amazon River, on the edge of the Peruvian rainforest. “Most people are self-employed or have several jobs to try and make ends meet,” Vazquez-Prokopec says. Common occupations of poorer residents include driving makeshift taxis or selling produce at one of the multiple open markets in the city.

Some previous studies in other parts of the world have looked at cell phone data to track and model human movements. The data is limited, however, due to issues of antenna density, restricted information from cell-phone carriers and the fact that some people do not have cell phones.

For the Iquitos study the researchers outfitted 582 residents with an i-GotU GPS device, which is ordinarily used as a photo-tagging tool for hikers. The i-GotU was selected for the study because it is small (about the size of a thumb drive), waterproof, relatively affordable, has a large memory and long battery life, and is password protected.

Each study participant wore one of the devices like a necklace as they went about their daily routines during a two-week period. The devices were programmed to capture location data every 2.5 minutes, from 5 am until midnight.

About 70 percent of the world's 3.3 billion city dwellers live in resource-poor environments. Aerial view of Iquitos by Viault / Wikipedia Commons.

The study yielded more than 2 million raw GPS positions, with an error margin of just four meters, tagged with date and time. The researchers used a data-reduction algorithm to calculate the average number of locations visited each day for the study participants as a whole, and by age groups, ranging from 7 years old to 60.

The results show that the participants visited an average of six locations per day overall. People in the peak working age group of 36 to 45 visited an average of nine locations daily.

“The more random your movements are, the more chances you have to pass a pathogen like the flu,” Vazquez-Prokopec says, explaining the 20 percent higher transmission risk, compared to a developed city.

While the Iquitos study represents just one city in the developing world, the researchers hope that the fine-scale, spatial-temporal data they have gathered will help fill the knowledge gap on human mobility in similar cities.

About 70 percent of the world’s 3.3 billion city dwellers live in resource-poor urban environments. “Uncovering the basic mechanisms governing complex human behaviors in these environments is paramount for developing better infrastructure, fostering economic development and responding to infectious disease threats,” Vazquez-Prokopec says.

The Emory research team also included Uriel Kitron, chair of the department of Environmental Studies, and post-doctoral fellow Donal Bisanzio. The study is part of a larger, ongoing disease ecology project centered in Iquitos that also includes scientists from the University of California, Davis, the U.S. Navy, the University of Iowa, Tulane University, San Diego State and researchers in Peru.

Related:
How the dengue virus makes a home in the city
Disease trackers take aim at dengue fever

No comments:

Post a Comment