Predicting and Tracking the Swine Flu Virus Part 1

Predicting and Tracking the Swine Flu Virus
By Susan Smith

Geospatial technology has become an integral part of predicting and tracking of infectious diseases such as the Swine Flu H1N1 virus pandemic which dominates our news today. According to Seth Wiafe and Bill Davenhall, ESRI Health and Human Services Solutions, disease surveillance using GIS provides public health officials with the information needed to detect and manage disease outbreaks. They say that “successful disease surveillance activities require standardized methodology, appropriate tools for prompt data collection, accurate synthesis of the data, continuity over time, and—most important—timely dissemination of the resulting information to health officials and, as appropriate, the public.”

In each one of these cases, early detection can make all the difference in being able to take effective action to contain a major outbreak of a disease. The ways that GIS can help in outbreak response are through disseminating information about threats of emerging infectious diseases and enhancing decision support at all levels including local, regional and national. Additionally, GIS can be useful in planning disease surveillance activities and predicting outcomes before financial commitments of public health intervention are made, and in allocating resources.

Mapping and spatial analysis of the spread and patterns of disease also reveal other information such as trends and interrelationships and can be used with georeferenced epidemiological data that take into account demographic, environment, hydrologic and vegetation information. The fact that maps can be dynamically updated as new information is received render them an invaluable resource, both for those working in surveillance mode and the public.

Predicting the Spread of the Virus

Proving that great minds think alike, a computer model at Northwestern University is predicting the Swine Flu epidemic’s future. Its forecast suggests approximately 2,000 cases by the end of this month, predominately in New York, Los Angeles, Miami and Houston. Simultaneously, informatics professor Alessandro Vespignani at Indiana University has come up with strikingly similar data with GLEaM, a Global Epidemic and Mobility modeler that integrates sociodemographic and population mobility data in spatially structured stochastic disease models to simulate the spread of epidemics at the worldwide scale.

Both these models enlist the aid of supercomputers to predict the spread of the viruses. Indiana University’s algorithms are different from those of the computer model at Northwestern. Dirk Brockmann, engineering professor lead on the epidemic modeling team at the Northwestern Institute on Complex Systems, said that the number of cases is growing faster than either of the two models had predicted.

Brockmann’s simulation employs two large datasets: air traffic and commuter traffic patterns for the entire country, and the data from the Web site, Where’s George? Where’s George? was dreamed up by programmer, Hank Eskin, over ten years ago. Eskin marked each dollar bill he received with a note asking the next owner to enter its serial number and zip code into his Web site, in order to track how far and fast the bills would travel. By 2006, the site had tracked the histories of 100 million bills. Brockmann uses this Web site because it relies on face-to-face transactions, which are what spreads influenza.

It takes two days to develop a simulation using either of the two models, according to Brockmann and Vespignani.

User login