Tracking the flu with George Washington

May 2, 2009


Northwestern professor Dirk Brockmann thinks he can track the spread of H1N1 flu and other future epidemics, thanks in part to the humble $1 bill.

“We’re looking at how people travel in the United States and Europe and trying to find a theory behind human traffic,” Brockmann said. “Then we can unravel the structures within these networks and explain them.”

Modern-day flu epidemics are fueled in part by the ubiquity of global air travel. Local commuting and mobility patterns can be equally important, but more difficult to analyze.

In 2006 Brockmann turned to WheresGeorge.com, a website that tracks dollar bills embossed with unique serial numbers, using money as a proxy for human travel. By calculating the odds that a given dollar bill will stay within a 10 kilometer radius, he created an epidemiological model that can be used to predict the course of the flu outbreak.

Brockmann explains his work on the current H1N1 outbreak in the following video

From Northwestern News on Vimeo.

Below you can see Brockmann’s worst-case projection for spread of the H1N1 flu virus in the United States, on a county-by-county basis. The model predicts up to 1,700 cases of swine flu for the entire United States four weeks from now, though Brockmann notes on his website: “The actual case numbers are expected to be smaller as mitigation strategies and containment efforts become effective.”

[QUICKTIME http://rocs.northwestern.edu/projects/swine_flu/index_assets/WGc5MEX_20090430_stoch.mp4 480 320]

His site also has worst-case projections for New York, Los Angeles, Miami,  Houston, Dallas and Chicago.

***
More:

Read about WheresGeorge.com at Associated Content.

One comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

Has anyone else noticed how completely inaccurate these predictions have been? As of 3 p.m. on May 6, Prof. Brockmann was still projecting only 20-80 cases in the Chicago area by May 14. Yet as of today, there are 102 cases in Chicago and another 59 in Cook County (at least some of which fall under the “Chicago area” heading.

When the outbreak initially occurred, his model predicted no more than 100 cases in the Chicago area by May 30; no more than 30 cases by May 14. AND the model emphasizes that his predictions are worst-case scenarios in which no measures are taken to prevent the spread of the disease.

Hopefully this experience will aid researchers in creating a better model for the prediction of something of this nature.

Posted by R.J. | Report as abusive