"I think the way to tell, obviously, is to compare it with actual surveillance data, which they should be able to pick up locally," said Stephen Morse, an epidemiologist at Columbia University's Mailman School of Public Health. "Whether searching for the flu is meaningful or not can only be compared with another gold standard method."
Morse said the system is open to a few problems.
"I think the approach sounds like a reasonable one, but I can think of several reasons why people might want to search for information on the flu," he said.
For example, Morse noted, as a flu researcher, he searches for flu information constantly, despite not being ill. More prevalent problems could accompany an increase in flu-related search terms because of a media report about the flu or, at this point, the threat of avian influenza.
Others have similar concerns.
"It's certainly intriguing and outside-of-the-box," said Dr. Lisa Jackson, an infectious disease specialist with the Group Health Center for Health Studies in Seattle. "The difficulty is that the term, flu, is not just used for influenza. A lot of people think of flu as a gastrointestinal issue."
She cited the rhinovirus, which causes stomach flu, along with the norovirus and rhinovirus, which cause the common cold -- which many mistake for flu -- as possible confounders of Google's new system.
"I think a big problem is that the term, flu, is used for lots of different things, so right away you have an issue with that," Jackson said. "I'm not sure how well [Google Flu Trends] will translate into helping us identify areas of influenza activity."
In response to concerns like this, Ginsberg said that in researching the new site, the data that is gathered from search engine queries is compared against past search trends and physician data. In the past, the trends in searching for flu have lined up well with actual outbreaks.
"What we have done is actually looked at four or five years of all of the data collected by the CDC's network of physicians. ... We compared that to the aggregated search queries on the flu over the past five years. We realized that the correlation tends to be very, very tight."
It remains unclear, however, how the information will be affected by the possibility of manipulation on the Internet by people hoping to alter results.
Still, while they wait to see the results of the new system, researchers say that unconventional methods like this have had some success in the past.
Jackson cited an outbreak of diarrhea several years ago in Milwaukee that was caused by a faulty septic system. She noted that the outbreak was caught because pharmacists in the area noticed that the diarrhea drug Imodium and toilet paper were flying off the shelves.
Morse noted that another Web site with a similar concept to Google Flu Trends, called "Who is Sick?" asks sick individuals to describe their symptoms. They can then observe if others in their area have similar illnesses.
Morse also indicated that better surveillance is certainly needed for figuring out where outbreaks of flu occur.
"Determining when a flu outbreak begins is difficult because there's a tendency not to report it," he said.