Title: Functional data analysis characterizes the shapes of the COVID-19 epidemic in Italy
Authors: Marzia Cremona - Université Laval (Canada)
Francesca Chiaromonte - The Pennsylvania State University (United States) [presenting]
Tobia Boschi - The Pennsylvania State University (United States)
Jacopo DiIorio - Sant Anna School of Advanced Studies (Italy)
Lorenzo Testa - Sant Anna School of Advanced Studies (Italy)
Abstract: COVID-19 mortality across 20 Italian regions is investigated, as well as its association with mobility, positivity, socio-demographic, infrastructural and environmental covariates. Notwithstanding limitations in accuracy and resolution of publicly available data, we pinpoint significant trends exploiting information in curves and shapes with functional data analysis. For the first epidemic wave (Feb-May 2020), we identify two starkly different patterns; an exponential one unfolding in Lombardia and the worst-hit areas of the north, and a milder, flat(tened) one in the rest of the country - including Veneto, where aggressive testing was implemented. We find that mobility and positivity predict mortality, also when controlling for relevant covariates. Among the latter, primary care appears to mitigate mortality, and contacts in hospitals, schools and workplaces aggravate it. Extending our analyses to the second epidemic wave (Oct 2020-Feb 2021) we find differences in mobility restrictions compared to the first, but we confirm a strong role for mobility and a marked heterogeneity in mortality patterns across the country. FDA techniques could capture additional signals if applied to richer data.