March 13, 2010 | 1
MIAMI—Before a government reports a disease outbreak, cases must usually be counted, verified and assessed—a process that can take days, weeks or months.
This delay creates a tension, however, as faster responses are the best hope for keeping a local outbreak from becoming an epidemic or even a pandemic. Overall, bureaucratic diligence and even political reluctance can keep a lid on emerging infections for too long, experts remarked here Friday at the 14th annual International Congress on Infectious Diseases.
Following trends in law enforcement and the media, public health workers more recently are calling on individual clinicians, lay people and local residents—not just officially validated data—to keep them informed of outbreaks in an effort to do better at staying ahead of health trends before they turn into big problems.
For years, doctors and experts in the field had been "very uncomfortable" about using anecdotal or even news reports of infections until the information could be corroborated with reviews from formal sources such as governments or the World Health Organization (WHO), said Marjorie Pollack of ProMED-mail, a public online reporting system for emerging diseases.
Now, however, with people, plants and animals traveling quickly—and information even more so—it doesn’t make sense to block out those sources of information, noted Larry Madoff, also of ProMED. "Monitoring informal sources of information—or rumors—can be an important source," he said. And that’s just what his group aims to do. By searching wire services, blogs and notifications from their 40,000 subscribers, the team assembles a daily list of reports from around the globe about infections that might affect humans—whether that is a crop contagion that could mean economic hardship, an animal disease that could result in a food shortage or a human (or zoonotic) illness.
When an infection is first reported, "we don’t know who needs to know," Madoff said, explaining why ProMED publishes reports on everything from maize wilt in Minnesota to measles in the Philippines to subscribers in 185 countries.
In February 2003, for example, ProMED received an email from a subscriber asking for any news about pneumonia circulating in Guangdong province in China. ProMED hadn’t but looked into it anyway. That tip turned out to be the start of SARS (severe acute respiratory syndrome), which cropped up in Canada within weeks. And even that continental jump was reported to them early. The day before the WHO issued an official alert about the disease, a doctor in Ontario, Canada, wrote in to ProMED to report seeing similar symptoms in a patient who had returned from Hong Kong.
As pinpointing the rapid dissemination of diseases has become increasingly important, another group, HealthMap, has partnered with ProMED and other groups to assemble geographic-based disease reports. Their program scans "tens of thousands of sites per hour" searching for clues about emerging infections, explained John Brownstein, of HealthMap. The company uses a five-way spam filter to sort through duplicate and questionable reports to come up with about 600 items daily to populate its interactive map and lists, he said.
The group also encourages users to report information from their neck of the woods, creating what Brownstein called "participatory epidemiology." He said he thinks of HealthMap "almost like the Wikipedia of emerging infectious diseases." Last year the group launched a smartphone application called Outbreaks Near Me, which got more than 100,000 downloads in the first month it was available, he noted.
Although HealthMap and ProMED don’t purport to replace traditional sources of disease information such as the U.S. Centers for Disease Control and Prevention (CDC), Brownstein and Madoff noted that they hope their services lend more speed and nuance to the overall understanding of outbreaks. Government sources, such as the CDC’s Yellow Book for travelers, for example, often provide blanket disease risks for a whole country where diseases like dengue fever might actually be endemic in some areas but remain absent in others. But by gathering hyper-local data, Brownstein noted, they can help "to redraw the boundary of risk."
Other groups are homing in on specific infections to try to get a detailed, statistically sound picture of spread across time and geography through seemingly superficial online data. Google Flu Trends launched its U.S. seasonal influenza tracker in November 2008, which tracked Web search terms that had historically matched up with spikes in previous annual CDC flu figures. But, said Corrie Conrad of the Trends team, they were still ramping up when they got a tip in April 2009 saying they should take a look at flu search trends south of the border. Sure enough, in Mexico the team found strangely seasonal search terms cropping up in the springtime. So they assembled a new model and started tracking H1N1 as it spread in Mexico and the U.S.
Google Flu Trends, however, has been limited to what the team set out to do: find the trends in flu incidences via Google searches. When asked if their methodologies could be expanded to other infectious diseases, say malaria in parts of Africa, Conrad said it wasn’t likely. The group’s accuracy, Conrad said, depended in large part on a wealth of historical health data and heft of Web searchers who are in the habit of looking for influenza info online. In many lesser-developed, rural areas where other infectious diseases are common, Internet coverage is likely to be sparse and health reporting spottier. But with the spread of the Web and mobile devices, soon more people might be able to log on and add to the emerging trend of real-time infection tracking where it is needed most: everywhere.
Image courtesy of iStockphoto/JuSun