(a–c) Distribution of the population in tiles for the cities under analysis: each tile is coloured according to the number of individuals (see colour bar on the left-hand side). (d) Schematic representation of a tile, with household contacts in blue and social contacts in orange–some of the latter connect to tiles not visible in the diagram. The dashed grey lines indicate the boundaries of the tile.
Image source: Mazza F, Brambilla M, Piccardi C, Pierri F, Proceedings of the Royal Society A (CC BY 4.0)
Using contact networks based on real data and a mathematical model that distinguishes between citizens ‘adhering’ and ‘not adhering’ to public health measures, the researchers simulated the spread of an epidemic in three Italian cities: Turin, Milan and Palermo.
The simulations showed that even a small proportion of non-adherents can substantially increase the infection peak and hasten the time of infection, straining healthcare facilities. The effect is particularly evident when the transmissibility of the disease is moderate. Moreover, when people who ignore preventive measures are concentrated in certain areas, local hotspots can form, making it more difficult to contain the epidemic.
‘We observed that although the epidemic followed similar overall dynamics in the cities we studied,’ explains Francesco Pierri, a researcher in the Data Science Lab and coordinator of the study, ‘the geographical distribution of non-adherent behaviour significantly alters local contagion trajectories, leading to marked differences between city districts.’
This work emphasises the need to monitor the distribution of preventive behaviour and adapt public health strategies to the nature of individual urban contexts, thereby making public health interventions more effective.
The research is part of the CODE — Coupling Opinion Dynamics with Epidemics project, financed by NRRP Mission 4 ‘Education and Research’ – Projects of National Interest (PRIN 2022 NRRP).
Source: Politecnico di Milano