Posts Tagged ‘h1n1’

GIS Applications in Epidemiology

Wednesday, December 30th, 2009

Thanks, JZ for the post

Applications of GIS aren’t new to epidemiology. Dating back to 1854, during a cholera outbreak in London’s Soho district, Dr. John Snow plotted the location of every individual case on a map and determined that they were distributed in a certain pattern that was linked to a contaminated water pump used by the local citizens. Now GIScience is being used to used to track the spatial distribution of all sorts of diseases.

H1N1 is a current epidemiological problem. Although H1N1 has been tracked since the outbreak, a lack of effective analysis tools (and countermeasures, of course) meant that the flu spread throughout the world within a few months. According to the latest update from the WHO, over 11,516 have died in the pandemic.

ESRI’s GIS is being used to track H1N1. According to ESRI’s own whitepaper, ‘GIS and Pandemic Influenza Planning and Response’, ESRI believes that geographic accuracy is essential in any infectious disease outbreak, and GIS applications can be critical in assessing risks, evaluating threats, tracking outbreaks, and ensuring the focused allocation of resources (e.g., vaccines and antivirals).

GISs tend to be rather static in their ability to model time. What is especially important is to be able to dynamically run a geospatial model of the outbreak. According to a recent article in Nature, agent-based modeling (ABM) can be used in modeling the disease’s possible spread, and designing policies for its mitigation. The ABM is basically an artificial society. Every person is represented by an autonomous software agent. Agents interact with each other; the computer tracks the agents’ health status as they interact in the virtual social network. Unlike classical epidemic modeling which based on differential equations, the ABM can simulate the complexity of social network. ABMs can be used to answer questions like, ‘What if a significant number of people refuse the H1N1 vaccine out of fear?’ ‘What is the best way to allocate the limited supplies of vaccines?’ or ‘How effective are school closures?’

A U.S. scale ABM (containing 300 million agents) can be run in approximately 10 minutes and can present the results on a map-based interface. Thus GIS and ABMs can provide the decision-makers a quick feedback on how their interventions work. As H1N1 moves through time and space and other possible pandemic influenza emerge in the future, GIS and ABM will play important roles in improving the efficiency of health agencies.