During a visit to the Universidad de Chile Center for Mathematical Modeling (CMM), Carlos Castillo-Chávez explained some of its advances in the modeling of disease and its spread.
The professor at Arizona State University and member of the U.S. President Barack Obama’s Committee on the National Medal of Science discussed several cases with a group of CMM researchers. The expert was invited by CMM and Universidad Federico Santa María.
The academic explained how models are used to determine the expansion of epidemics by considering design of cities, movement of people and sociological studies that describe citizens’ habits. Comparing that the population does under normal situation and what is done when precautions are taken.
He furthered the expansion of Ebola in 2000, analyzing how some control mechanisms as the cordon sanitaire -when the army blocked exits of cities- may have dissimilar effects: meanwhile in urban areas with well-implemented health systems can avoid expansion of diseases, in poor villages can cause more deaths due to scarce access to doctors and medications and increment of contacts between neighbors. For the professor, this last factor becomes more relevant to predictive models than routes followed by people in a city, which is a factor that had traditionally considered as the main. To understand this, Castillo-Chávez and his researchers used a Lagrangian method, where times of residence spent by individuals in different areas were more important than their movements through the urban area.
The expert in Mathematical Biology also showed an example of how technology and the different actions taken by government may affect predictive models of progress. It was the case of influenza in Mexico during 2010.
“Before, a disease moved like by a train; Today, it does like by an airplane”, he said. “Also medicines affect; if there was gets influenza again, it will no longer follow the same pattern of contagion.”
In the Mexican case, the closure of schools and the arrival of summer explained why the disease’s spread did not occur as models predicted, which implied changes in the algorithms used and a continuing need of revision.
