266515 Spatial distribution of pandemic influenza emergence: Development of a global logistic regression model using eco-geographic indicator variables

Tuesday, October 30, 2012 : 9:05 AM - 9:20 AM

Nita Madhav, MSPH , Research and Modeling, AIR Worldwide, Boston, MA
Benjamin Spaulding, PhD , Research and Modeling, AIR Worldwide, Boston, MA
Vladimir Ivanov, PhD , Research and Modeling, AIR Worldwide, Boston, MA
John Rowe , Research and Modeling, AIR Worldwide, Boston, MA
Introduction: There is great uncertainty regarding the location of the emergence of the next influenza pandemic. Methods: We used principal components analysis (PCA) and logistic regression modeling to create a global emergence probability map for pandemic influenza. A logistic regression model was fit using reported human influenza cases of zoonotic origin, and controls were sampled randomly from areas where cases had not been reported. Eco-geographic indicator variables that were used included poultry and pig density, human population density, migratory waterfowl species richness, land use/land cover data, and reported cases of influenza in animals. These data were transformed using PCA and included in the logistic regression model. Results: The first four principal components explained 83% of the variance among the eco-geographic variables. PCA factor loadings were as follows: 1) human contact with domestic livestock; 2) presence of wild avian reservoirs; 3) contact between animal influenza cases and domestic livestock; and 4) contact between animal influenza cases and human populations. The logistic regression model with the best fit included the first four principal components and an interaction term between the first and third principal component. All model coefficients were significant, with p-values <0.001. The highest probability of emergence was found to coincide with known areas of avian influenza circulation in Southeast Asia and Africa. Increased emergence probability was also found in North America, including the areas of Mexico where the 2009 H1N1 pandemic is thought to have emerged. Conclusion: Surveillance and preventive measures could be targeted towards areas found to have increased probability of influenza pandemic emergence.

Learning Areas:
Environmental health sciences
Epidemiology
Public health biology

Learning Objectives:
Describe six important eco-geographic variables associated with pandemic influenza emergence Identify areas with increased likelihood of pandemic influenza emergence

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I am qualified to present because I am the scientific lead for the pandemic influenza modeling effort at my organization. I have a BS in Ecology and Evolutionary Biology from Yale University, and an MSPH in Epidemiology from Emory University.
Any relevant financial relationships? Yes

Name of Organization Clinical/Research Area Type of relationship
AIR Worldwide Catastrophe Modeling Employment (includes retainer)

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.