Seasonal Flu Waves in the US
Last week I read an extremely interesting article in (super awesome open access) PLoS One, “Seasonal Synchronization of Influenza in the United States Older Adult Population.” Written by researchers out of Tufts University School of Medicine, this article looks at CMS data about hospitalization data from thirteen flu seasons (1991 through 2004) in order to learn when and where flu happens. The results paint a vivid picture of what happens during a typical flu season here in the States; a picture which can help inform traditional public health response to the annual epidemic that takes tens of thousands of lives every year.
As you can see in the picture below, the peak timing of seasonal flu spreads across the country, starting the West (e.g., California, Oregon, Nevada, Utah and Colorado) and progressing eastward through the Midwest, the South, the Mid-Altantic states and finally into New England (where Maine, New Hampshire and Vermont experience the flu peak last). The entire process, from the start of the seasonal peaks through the end, took just four weeks.
This has huge implications for public health preparedness specifically in how it informs public health response in general. Every year, public health departments and hospitals across the country are forced to respond to a tremendous influx of patients with varying levels of morbidity and mortality. Many act under the axiom that flu is different everywhere, state to state, city to city, neighborhood to neighborhood. And at any point in time, that might be the case. But we now see that, at least at the state level, there is some rhyme and reason to the spread of the disease.
Think of what this knowledge could change. Knowing that flu has peaked in the western states could very well cause inform decisions made by East Coast hospitals. Think of what a four-week heads up on a surge of flu cases could do for flu vaccine outreach efforts. Beyond flu, this could be the first in a series of article examining the chronological spread of other yearly epidemics like rhinovirus, RSV or norovirus. I fully expect to see this type of analysis for H1N1 influenza peaks over time and space. When it was everywhere, did it peak everywhere at the same time or did it follow the same pattern seen here?
In addition to the chronological peaking data, the authors noted another phenomenon. Seasons that peaked early tended to be more intense, with more cases throughout the full season. What an amazing early warning system.
I’ll admit that this study is built from a relatively small sample, but the implications of such studies are great and could help to mount an effective response to a seasonal scourge. I look forward to studies that expand greatly on this idea.
The press release, with the above picture, can be found here.