The role of swabs in modeling the COVID-19 outbreak in the most affected regions of Italy

TL;DR

This paper introduces a nonlinear asymmetric diffusion model incorporating daily swab data to better understand COVID-19 case fluctuations and outbreak dynamics in Italy's most affected regions.

Contribution

It presents a novel diffusion model that accounts for testing variations, improving outbreak modeling and estimation of total cases compared to existing models.

Findings

Model effectively captures daily fluctuations influenced by testing changes.

Incorporating swab data improves outbreak trend accuracy.

Model provides reliable estimates of total confirmed cases.

Abstract

The daily fluctuations in the released number of Covid-19 cases played a big role both at the beginning and in the most critical weeks of the outbreak, when local authorities in Italy had to decide whether to impose a lockdown and at which level. Public opinion was focused on this information as well, to understand how quickly the epidemic was spreading. When an increase/decrease was communicated, especially a large one, it was not easy to understand if it was due to a change in the epidemic evolution or if it was a fluctuation due to other reasons, such as an increase in the number of swabs or a delay in the swab processing. In this work, we propose a nonlinear asymmetric diffusion model, which includes information on the daily number of swabs, to describe daily fluctuations in the number of confirmed cases in addition to the main trend of the outbreak evolution. The class of diffusion models has been selected to develop our proposal, as it also allows estimation of the total number of confirmed cases at the end of the outbreak. The proposed model is compared to six existing models, among which the logistic and the SIRD models are used as benchmarks, in the five most affected Italian regions.