The Evolution of Covid-19 in Italy after the Spring of 2020: An Unpredicted Summer Respite Followed by a Second Wave

Abstract

The coronavirus (COVID-19) pandemic was particularly invasive in Italy during the period between March and late April 2020, then decreased in both the number of infections and in the seriousness of the illness throughout the summer of 2020. In this work, we measure the severity of the disease by the ratio of Intensive Care Units (ICU) spaces occupied by COVID-19 patients and the number of Active Cases (AC) each month from April to October 2020. We also use the ratio of the number of Deaths (D) to the number of Active Cases. What clearly emerges, from rigorous statistical analysis, is a progressive decrease in both ratios until August, indicating progressive mitigation of the disease. This is particularly evident when comparing March-April with July-August; during the summer period the two ratios became roughly 18 times lower. We test such sharp decreases against possible bias in counting active cases and we confirm their statistical significance. We then interpret such evidence in terms of the well-known seasonality of the human immune system and the virus-inactivating effect of stronger UV rays in the summer. Both ratios, however, increased again in October, as ICU/AC began to increase in September 2020. These ratios and the exponential growth of infections in October indicate that the virus-if not contained by strict measures-will lead to unsustainable challenges for the Italian health system in the winter of 2020-2021.

Keywords: Covid-19 in Italy; autumn sharp worsening of epidemics; summer mitigation.

Figure 1

(a) Ratio between ICU occupation and active cases in each month, from April to October. Error bars are as in Table 1 (statistical errors at 95% probability level, i.e., two standard deviations). (b) Ratio between deaths and active cases in each month, from April to October. Error bars are as in Table 1 (statistical errors at 95% probability level, i.e., two standard deviations).

Figure 2

(a) Relative ratios of ICU/AC in April, with respect to the following months May to August. Error bars are as in Table 1 (statistical errors at 95% probability level, i.e., two standard deviations). (b) Relative ratios of D/AC in April, with respect to the following months May to August. Error bars are as in Table 1 (statistical errors at 95% probability level, i.e., two standard deviations).

Figure 3

(a) From left to right: climatologies of the irradiance at 307.5 nm, 322.5 nm, and the UV index at two stations in Italy: Aosta, Northern Italy, Rome, Central Italy, in the period 2006–2015. Shaded intervals correspond to the standard deviation of the climatological values and the average ratios (redrawn from [24]). (b) Comparison of the official daily infection curve in Italy from March 1st to mid-October (red line) with the average (2006–2015) UV intensity curve recorded at Rome station.

Figure 4

(a) Ratio between hospitalized Covid-19 patients and total number of active cases, in Italy; (b) ratio between ICU occupation and hospitalized patients. The considered period is February 24th (indicated as day 1) to October 23rd.

Figure 5

Extrapolated exponential curves for daily infection data since August 15th to October 23rd (red circles). The best fitting exponential curve for all data is shown by the black line; red line shows the best fitting curve using only data since September 15th to October 23rd.