Predicting the Impact of COVID-19 on the Emergency Departments in Lombardy, Italy

• Activity Number: 41 - Storytelling on COVID-19 Impact Using Experts' Prior Knowledge and Data from Social Media, Official Clinical Data, Digital Phenotype from Smartphones' Raw Sensor Data, and Emergency Departments
• Type: Invited
• Date/Time: Sunday, August 8, 2021 : 3:30 PM to 5:20 PM
• Sponsor: International Society for Bayesian Analysis (ISBA)
• Author(s): Antonietta Mira and Giulia Ghilardi and Greta Carrara and Angela Andreella* and Spyros Balafas and Fabrizio Ruggeri and Ernst C Wit and Livio Finos and Guido Bertolini and Giovanni Nattino
• Companies: Università della Svizzera italiana and University of Insubria and Istituto di Ricerche Farmacologiche Mario Negri IRCCS and Istituto di Ricerche Farmacologiche Mario Negri IRCCS and University of Insubria and University of Insubria and CNR IMATI and Universita della Svizzera italiana and University of Padova and Laboratory of clinical epidemiology and Istituto di Ricerche Farmacologiche Mario Negri IRCCS
• Keywords: Covid-19; Emergency departments data; Emergency call data; Predictive models; Decision support systems

Abstract

Italy, and in particular the Lombardy region, was among the first countries outside of Asia to report cases of COVID-19. The Lombardy region relies on the emergency medical service called Agenzia Regionale Emergenza Urgenza (AREU). It coordinates the intra- and inter-regional non-hospital emergency network and the European emergency number service. Therefore, AREU must deal with daily and seasonal variations of call volume. Many factors can describe this call volume across time beyond the annual trend, such as weather circumstances and epidemiological factors. Factors related to the day of the week, time of the day, seasonal and yearly variations that characterize the time series pattern must also be considered. In addition, the number and type of the emergency calls changed dramatically during and after the COVID-19 epidemic peak. Statistical modeling is essential for AREU to predict incoming calls and how many of these turn into events, i.e., dispatch of transport and equipment until the rescue is completed. In this talk, we present a Generalized Additive Model able to predict the number of events during the COVID-19 pandemic with an error compatible with the AREU requests.