Spatio-temporal resource mapping for intensive care units at regional level for COVID-19 emergency in Italy

COVID-19 is a worldwide emergency since it has rapidly spread from China to almost all the countries worldwide. Italy has been one of the most affected countries after China. North Italian regions, such as Lombardia and Veneto, had an abnormally large number of cases. COVID-19 patients management requires availability of sufficiently large number of Intensive Care Units (ICUs) beds. Resources shortening is a critical issue when the number of COVID-19 severe cases are higher than the available resources. This is also the case at a regional scale. We analysed Italian data at regional level with the aim to: (i) support health and government decision-makers in gathering rapid and efficient decisions on increasing health structures capacities (in terms of ICU slots) and (ii) define a geographic model to plan emergency and future COVID-19 patients management using reallocating them among health structures. Finally, we retain that the here proposed model can be also used in other countries. COVID-19 [1] is caused by the SARS-CoV-2 virus and belongs to the Coronaviridæ family, which contains many other viruses. Only seven of which are known to be responsible for human diseases, e.g., 229E, NL63, OC43, HKU1, MERS-CoV, SARS-CoV, and SARS-CoV-2 [2,3]. The virus diffused with a surprisingly fast pace, and in one month putting under stress the healthcare resources worldwide, starting from China. Italy was the first European country affected by the virus. The high spreading rate and the absence of tailored therapies and vaccines determine a relatively high mortality rate that has been controlled by blocking the virus spreading with severe mobility restrictions to the people of the infected regions [3]. By the end of March, while the situation in China seems to be under control, the virus is rapidly growing in other countries [4,5]. With different time scales, other countries such as the USA, France, Spain and North Europe reacted by implementing containment measures. The virus has an initial exponential diffusion which requires: (i) home quarantine for low symptoms, (ii) hospitalisation for part of them and, (iii) hospitalisation in ICUs requiring respiratory support for severe ones. In some cases, COVID-19 causes severe pneumonia, especially in the presence of co-morbidities [6], thus patients need hospitalization in ICU where respiratory support (such as mechanical ventilators) are required to keep them alive [7]. We focus on a disease diffusion model which helps predicting ICU resources, for the Italian emergency. The model is general enough to foresee its adoption also in other countries. It also scales well at a regional or sub-regional level.