How flu virus trackers keep you healthy
By keeping an eye on where the flu appears, experts can help doctors better diagnose symptoms and vaccine makers to improve next year's shots.
Fri, Sep 20, 2013 at 12:36 PM
Feel a tickle in your throat? A bit feverish? As flu season approaches, you might want to see your doctor. But another kind of health professional will take an interest, too: call them the flu watchers.
With influenza's yearly arrival imminent, a vast network of epidemiological surveillance systems is springing back to life. Officials in all 50 states and within the Centers for Disease Control and Prevention (CDC) will be working to discover when, where and with what severity flu strikes in 2013.
Thanks to the virus' ever-present pandemic threat — and its unique behavior — influenza warrants this yearly obsession. The surveillance helps doctors keep people healthy, and provides the first warning signs of dangerous new strains.
What sets flu apart
To put it bluntly, influenza is a weird virus. Unlike similar illnesses, the flu makes wildly unpredictable movements every year, said Dr. Joseph Bresee, chief of the epidemiology and prevention branch of the CDC's Influenza Division. Rotavirus, for example, always starts in the Southwest United States and spreads northeastward. You can plot its movement from region to region, like a wave spreading across a lake.
But flu can start anywhere in the country. And as it moves from area to area, it follows no discernable pattern. "Flu is unique," Bresee said. "We know we'll get it every year, but where it will go and how severe it will be is impossible to predict."
The peak of flu season can occur anytime between November and March. The season can start in New York City, then jump to the Southwest — or follow any number of other routes. Influenza epidemiologists have a saying to describe this unpredictability, Bresee said: "If you've seen one influenza epidemic, you've seen one influenza epidemic." [6 Flu Vaccine Myths]
The virus itself is also pretty slippery. Known for mutating at a particularly fast rate, even for a virus, the influenza strain that strikes this year may differ from last year's. That means the vaccine must also change each season, Bresee said. "The flu vaccine is the only vaccine that changes every year," he said.
Tracking a serious illness
The changeability of the flu virus matters especially because of the severity of influenza, which causes about 200,000 hospitalizations in the United States on average each year. Over the past few decades, the number of yearly deaths in the U.S. from the flu has ranged from 3,000 to 49,000.
And when new strains of flu jump from animals to humans, they raise the potential for widespread, severe illness, as happened with the H7N9 strain of avian influenza in China this spring, and the 2009 "swine flu" epidemic that started in North America.
During normal, seasonal outbreaks and when novel viruses appear, a number of systems work together to give public health officials a better handle on the unpredictable virus. The first indications of flu's arrival usually come from doctors' offices. Thousands of local physicians file weekly reports to the CDC about the number of patients arriving with flu-like symptoms. State officials also gather reports on the number of deaths and hospitalizations from influenza, revealing the season's severity.
Once reports of flu-like symptoms pass a 2 percent threshold -- meaning that 2 percent of patients coming in to see their doctors are doing so because they have flu symptoms -- officials can say that flu has likely arrived, said Angie Maxted, an influenza epidemiologist at the New York Department of Health.
And that knowledge makes an immediate difference in how doctors treat patients, said Dr. Jonathan Temte, a flu-management expert at the University of Wisconsin-Madison. With flu properly heralded, doctors in the area can quickly diagnose sore throats and coughs as flu cases, prescribing anti-flu medicines.
It would be a waste of resources and money to use flu medicines for every case of flu-like symptoms — because a sore throat in July is almost certainly just a cold. With epidemiological surveillance, doctors have a better chance of treating only the patients who really need it. "Once you know influenza is in your area, and what it looks like, you can be more precise in diagnosing and treating it," Temte said.
State laboratories also test patient samples that doctors send in, to determine the types and sub-types of influenza viruses circulating. This helps officials construct the next vaccine, and, perhaps more important, spot potentially dangerous new strains. Since the "H5N1" avian influenza outbreak of 2006-2007, the CDC has required all labs to report any new or unidentified influenza viruses they see. That protocol identified the first two cases of swine flu in the United States in 2009, when a San Diego military lab found the virus in two children.
"That's why it's important to have samples typed and sub-typed," Maxted said. "It's possible that a new, pandemic strain could arise anywhere."
A flu forecast?
In the future, a new form of surveillance could take flu epidemiology into the business of prediction, not simply tracking. Last year, researchers at Columbia University and the National Center for Atmospheric Research devised a computer program to forecast the flu, based on methods used in weather forecasting.
The technology employs a flu-tracking method developed by Google. The technology company worked with the CDC to show that monitoring searches for flu-related terms could give an accurate estimate of where flu is occurring. That provides, for the first time, the kind of real-time flu data that could make prediction possible, said Jeffrey Shaman, an assistant professor of environmental health sciences at Columbia, who led the research. (There is at least a six-day lag time with traditional tracking methods, Shaman said.)
Weather-prediction models continually add in data about the current state of the weather system; with Google's flu trends, epidemiologists might finally have the equivalent of current state data for influenza, Shaman said.
The researchers just completed a real-time test drive of the model, and are set to publish those results in the next few months. If their predictive system becomes widely adopted, flu trackers could put out the call to vaccinate, start diagnosing flu symptoms, and even close schools before flu season strikes any particular region, he said.
"If you make a prediction four to five weeks in advance, people still have time to get vaccinated," Shaman said."
Follow Michael Dhar @mid1980. Follow us @livescience, Facebook & Google+. Original article on LiveScience.
Related on LiveScience and MNN: