"Using mathematical models to analyze complex epidemic data and gain key insights on epidemic dynamics and control"
by Connor Bender
FPJ hosted a seminar at the National Institute of Infectious Diseases (NIID) with Dr. Simon Cauchemez of the mathematical modeling of infectious diseases unit at the Institute Pasteur in Paris. Dr. Cauchemez spoke about the challenges that may be associated with analyzing epidemic data and discussed how mathematical modeling can help address said challenges. Specifically, he gave three example studies where mathematical modeling was used to gain insight into epidemic data and trends.
For the first example study, he showed how he used statistical modeling techniques to analyze trends on seasonal influenza in French households in 2004 and in Vietnamese households in 2014. Through mathematical modeling, information such as the generation time of the infection and the person-to-person probability of transmission rates in households were determined. For the second study, data were drawn and analyzed from a study on flu in Kelly Island in 1920. The 1920 data was then compared to the school outbreak study of 2009; it was concluded that the probability of transmission was the same but there were more children per class in 1920. In the last study in Bangladesh in 2016 shown through mathematical modeling, it was determined that women in a certain village had a higher transmission risk of the chikungunya virus than men.
After presenting the studies and findings, Dr. Cauchemez reiterated that while there are many challenges such as multiple routes of transmission and missing data/censorship, modeling has the tools and capacity to solve these challenges. He concluded that categories such as subclinical infections and severity of disease can be predicted through modeling but acknowledges that there is a huge need for collaboration between epidemiologists and data modelers in the future.