April 30, 2009

Modeling tool shows how public-health efforts can reduce flu's spread

"Visualizing how specific actions would help to control the spread of a flu pandemic is very valuable to public health officials," said David S. Ebert, a professor in Purdue's School of Electrical and Computer Engineering and director of the Purdue Visualization and Analytics Center, which leads visualization sciences for the Department of Homeland Security's Center of Excellence for Command, Control, and Interoperability. "The idea is to help officials make decisions in terms of when to use stockpiles of antiviral drugs, when to close schools and issue media alerts for social-distancing measures."

Based on published epidemiology literature and data from the 1918 flu pandemic, the tool, developed in collaboration with researchers at Pacific Northwest National Laboratories, simulates the flu's daily spread over a two-month period across Indiana. It compares how the disease progresses if steps are or aren't taken to control the spread and the impact of when the measures are implemented.

The tool, called PanViz, was first prepared for the Indiana State Department of Health, and researchers are working to provide it to other states. The visualization demonstrates how certain actions affect the availability of hospital beds, the percentage of ill people and the percentage of deaths.

"You can also visualize the disease's progress based on where it originates in the state and see the effect of population demographics," Ebert said. "For example, the spread would be slower in rural than urban areas, and so on."

The researchers have prepared a simple video to show a side-by-side comparison, displaying the changing, color-coded infection rate. The video, which assumes the first case was reported in Chicago, is available at https://pixel.ecn.purdue.edu:8080/~rmacieje/PanViz_Video.avi

"This tool does not provide a prediction, but a scientifically based simulation for showing the relative benefit of implementing different actions," Ebert said. "We do not have sufficient data to begin to predict the spread and impact, but given enough statistics, the model used in the tool will be more predictive. For instance, if we knew the expected spread rate and mortality from H1N1 infections in the United States, the tool would become more statistically accurate."

The research has been funded by the U.S. Department of Homeland Security through the Department's Center of Excellence program and the Indiana State Department of Health.

Social-distancing strategies to reduce contact between people include closing schools, reducing public gatherings, limiting personal interactions, working from home and voluntary quarantine. The tool is currently able to simulate the impact of using antiviral medications and implementing school closures and personal social distancing, such as proper cough etiquette, not shaking hands and restricting close interaction with others. Future research may expand the model to see how additional social-distancing actions, such as restricting public gatherings, canceling social events, and quarantines, would influence the spread.

Writer: Emil Venere, 765-494-4709, venere@purdue.edu

Source: David Ebert, 765-494-9064, ebertd@ecn.purdue.edu

Purdue News Service: (765) 494-2096; purduenews@purdue.edu

Note to Journalists: David Ebert is in Washington, D.C., this week. He is available for interviews before noon Thursday (April 30) and Thursday evening, Friday (May 1) before 10 a.m. and Friday from 2-3:30 p.m. He can be reached at 765-412-4433, ebertd@ecn.purdue.edu

To the News Service home page