A deep learning approach for Spatio-Temporal forecasting of new cases and new hospital admissions of COVID-19 spread in Reggio Emilia, Northern Italy.

BACKGROUND: Since February 2020, the COVID-19 epidemic has rapidly spread throughout Italy. Some studies showed an association of environmental factors, such as PM10, PM2.5, NO2, temperature, relative humidity, wind speed, solar radiation and mobility with the spread of the epidemic. In this work, we aimed to predict via Deep Learning the real-time transmission of SARS-CoV-2 in the province of Reggio Emilia, Northern Italy, in a grid with a small resolution (12 km × 12 km), including satellite information. METHODS: We focused on the Province of Reggio Emilia, which was severely hit by the first wave of the epidemic. The outcomes included new SARS-CoV-2 infections and COVID-19 hospital admissions. Pollution, meteorological and mobility data were analyzed. The spatial simulation domain included the Province of Reggio Emilia in a grid of 40 cells of (12 km)2. We implemented a ConvLSTM, which is a spatio-temporal deep learning approach, to perform a 7-day moving average to forecast the 7th day after. We used as training and validation set the new daily infections and hospital admissions from August 2020 to March 2021. Finally, we assessed the models in terms of Mean Absolute Error (MAE) compared with Mean Observed Value (MOV) and Root Mean Squared Error (RMSE) on data from April to September 2021. We tested the performance of different combinations of input variables to find the best forecast model. FINDINGS: Daily new cases of infection, mobility and wind speed resulted in being strongly predictive of new COVID-19 hospital admissions (MAE = 2.72 in the Province of Reggio Emilia; MAE = 0.62 in Reggio Emilia city), whereas daily new cases, mobility, solar radiation and PM2.5 turned out to be the best predictors to forecast new infections, with appropriate time lags. INTERPRETATION: ConvLSTM achieved good performances in forecasting new SARS-CoV-2 infections and new COVID-19 hospital admissions. The spatio-temporal representation allows borrowing strength from data neighboring to forecast at the level of the square cell (12 km)2, getting accurate predictions also at the county level, which is paramount to help optimise the real-time allocation of health care resources during an epidemic emergency.

Modeling Early Phases of COVID-19 Pandemic in Northern Italy and Its Implication for Outbreak Diffusion.

The COVID-19 pandemic has sparked an intense debate about the hidden factors underlying the dynamics of the outbreak. Several computational models have been proposed to inform effective social and healthcare strategies. Crucially, the predictive validity of these models often depends upon incorporating behavioral and social responses to infection. Among these tools, the analytic framework known as "dynamic causal modeling" (DCM) has been applied to the COVID-19 pandemic, shedding new light on the factors underlying the dynamics of the outbreak. We have applied DCM to data from northern Italian regions, the first areas in Europe to contend with the outbreak, and analyzed the predictive validity of the model and also its suitability in highlighting the hidden factors governing the pandemic diffusion. By taking into account data from the beginning of the pandemic, the model could faithfully predict the dynamics of outbreak diffusion varying from region to region. The DCM appears to be a reliable tool to investigate the mechanisms governing the spread of the SARS-CoV-2 to identify the containment and control strategies that could efficiently be used to counteract further waves of infection.

Associations between mortality from COVID-19 in two Italian regions and outdoor air pollution as assessed through tropospheric nitrogen dioxide.

After the appearance of COVID-19 in China last December 2019, Italy was the first European country to be severely affected by the outbreak. The first diagnosis in Italy was on February 20, 2020, followed by the establishment of a light and a tight lockdown on February 23 and on March 8, 2020, respectively. The virus spread rapidly, particularly in the North of the country in the 'Padan Plain' area, known as one of the most polluted regions in Europe. Air pollution has been recently hypothesized to enhance the clinical severity of SARS-CoV-2 infection, acting through adverse effects on immunity, induction of respiratory and other chronic disease, upregulation of viral receptor ACE-2, and possible pathogen transportation as a virus carrier. We investigated the association between air pollution and subsequent COVID-19 mortality rates within two Italian regions (Veneto and Emilia-Romagna). We estimated ground-level nitrogen dioxide through its tropospheric levels using data available from the Sentinel-5P satellites of the European Space Agency Copernicus Earth Observation Programme before the lockdown. We then examined COVID-19 mortality rates in relation to the nitrogen dioxide levels at three 14-day lag points after the lockdown, namely March 8, 22 and April 5, 2020. Using a multivariable negative binomial regression model, we found an association between nitrogen dioxide and COVID-19 mortality. Although ecological data provide only weak evidence, these findings indicate an association between air pollution levels and COVID-19 severity.

Lockdown timing and efficacy in controlling COVID-19 using mobile phone tracking.

BACKGROUND: Italy's severe COVID-19 outbreak was addressed by a lockdown that gradually increased in space, time and intensity. The effectiveness of the lockdown has not been precisely assessed with respect to the intensity of mobility restriction and the time until the outbreak receded. METHODS: We used processed mobile phone tracking data to measure mobility restriction, and related those data to the number of new SARS-CoV-2 positive cases detected on a daily base in the three most affected Italian regions, Lombardy, Veneto and Emilia-Romagna, from February 1 through April 6, 2020, when two subsequent lockdowns with increasing intensity were implemented by the Italian government. FINDINGS: During the study period, mobility restriction was inversely related to the daily number of newly diagnosed SARS-CoV-2 positive cases only after the second, more effective lockdown, with a peak in the curve of diagnosed cases of infection occurring 14 to 18 days from lockdown in the three regions and 9 to 25 days in the included provinces. An effective reduction in transmission must have occurred nearly immediately after the tighter lockdown, given the lag time of around 10 days from asymptomatic infection to diagnosis. The period from lockdown to peak was shorter in the areas with the highest prevalence of the infection. This effect was seen within slightly more than one week in the most severely affected areas. INTERPRETATION: It appears that the less rigid lockdown led to an insufficient decrease in mobility to reverse an outbreak such as COVID-19. With a tighter lockdown, mobility decreased enough to bring down transmission promptly below the level needed to sustain the epidemic. FUNDING: No funding sources have been used for this work.

Satellite-detected tropospheric nitrogen dioxide and spread of SARS-CoV-2 infection in Northern Italy.

Following the outbreak of Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2) last December 2019 in China, Italy was the first European country to be severely affected, with the first local case diagnosed on 20 February 2020. The virus spread quickly, particularly in the North of Italy, with three regions (Lombardy, Veneto and Emilia-Romagna) being the most severely affected. These three regions accounted for >80% of SARS-CoV-2 positive cases when the tight lockdown was established (March 8). These regions include one of Europe's areas of heaviest air pollution, the Po valley. Air pollution has been recently proposed as a possible risk factor of SARS-CoV-2 infection, due to its adverse effect on immunity and to the possibility that polluted air may even carry the virus. We investigated the association between air pollution and subsequent spread of the SARS-CoV-2 infection within these regions. We collected NO2 tropospheric levels using satellite data available at the European Space Agency before the lockdown. Using a multivariable restricted cubic spline regression model, we compared NO2 levels with SARS-CoV-2 infection prevalence rate at different time points after the lockdown, namely March 8, 22 and April 5, in the 28 provinces of Lombardy, Veneto and Emilia-Romagna. We found little association of NO2 levels with SARS-CoV-2 prevalence up to about 130 µmol/m2, while a positive association was evident at higher levels at each time point. Notwithstanding the limitations of the use of aggregated data, these findings lend some support to the hypothesis that high levels of air pollution may favor the spread of the SARS-CoV-2 infection.