To clarify both actual and official case fatality rates in comparison to those of other countries as well as the actual extent of the tested population, it is important to differentiate the case fatality rate from the actual fatality rate (infection fatality rate). The case fatality rate is the ratio of the number of deaths to the total number of positive tests, while the infection fatality rate is the ratio of the number of deaths to the total infected population. However, because the entire population (symptomatic and asymptomatic) cannot be tested with sufficient speed, it is necessary to rely on case fatality rates for epidemiology measures and policy planning. Moreover, as the epidemic escalates, the tested population represents the emergent cluster of patients who, by seeking care, turn to and contact national health systems.

At present, the COVID-19 case fatality rate in Italy is 12.80% [5]. In China, the case fatality rate shows a nonlinear trend, with a high rate at the start of the outbreak (17.3%) that decreased to 0.7% at the end of the emergency. China hit the infection curve plateau on February 20, 2020. At the time of writing, the Chinese case fatality rate is approximately 4.1% [5].

The current Italian case fatality rate is comparable with that of other European countries, such as Belgium (12.80%), the United Kingdom (12.70%), France (11.20%), and Spain (10.44%). To date, the worldwide case fatality rate is 6.40% [5]. In Germany, the case fatality rate is 2.7%; however, the Robert Koch Institute has stated that this rate is expected to grow greatly [7]. Furthermore, general misalignment with other European countries is observed for other German indices related to COVID-19, such as the distribution of deaths per age of infected people and the distribution of deaths per gender [8,9].

The Lombardy region is the region most affected by COVID-19 worldwide; data indicate that this region represents 38% of the total infections in Italy, with a regional case fatality rate of 18.08% [1]. No other countries have such unbalanced distribution and variability among their regions [5]; therefore, we can perform the statistical exercise of considering the Lombardy region value as an outlier (ie, a data point that differs significantly from the other regional observations). Hence, if we attempted to disaggregate the contribution of the Lombardy region proportionally to the national datum, we would obtain an Italian case fatality rate of 9.60%. The Italian case fatality rate value would decrease even further to 7.9% if data from the Emilia-Romagna region (the second most affected Italian region) were also excluded.

This statistical exercise stems from the consideration that the very first COVID-19 outbreak in Europe occurred in the Lombardy region, and the National Health System was realistically unprepared to face a pandemic.

Ultimately, it is important to note that the case fatality rate trend is not linear over time due to the number of total infected people depending on the daily number of new infections combined with the numbers of recovered and discharged patients. As shown in Figure 1, the trend line (ie, the 5-day moving average) of daily recoveries grows strongly over time, whereas the trend line for COVID-19 fatalities decreases with a different slope function.

Figure 2 also shows that the curve of total recovered/discharged patients grows much faster than the curve of total deaths. The difference in the slopes of the two curves could enable a reliable prediction of a decrease in the national case fatality rate at the end of the first wave of the epidemic.

Furthermore, as officially reported in the available open data [1], the occupancy of beds in COVID-19–dedicated intensive care units is steadily decreasing, with increasingly diverging curves between the number of new infections and the number of patients with severe symptoms that could lead to death in the near future. At the time of writing, there are more than 3000 patients with severe symptoms out of 104,000 total active positive cases.

Table 1 shows the COVID-19 case fatality rates and mortality rates (the ratio of deaths of patients who tested positive for COVID-19 to the total number of citizens) on April 14, 2020 among the industrialized countries in the G10: Belgium, Canada, France, Germany, Japan, Italy, the Netherlands, the United Kingdom, the United States, and Sweden [5]. Comparing the Italian data to those of the other G10 countries, the COVID-19 mortality in Italy is approximately 358 per 1 million inhabitants, which is not the highest value in the set. Belgium, for instance, has a mortality rate of 419 per 1 million inhabitants, and the rate in Spain is 409 per 1 million inhabitants. Following the model of the statistical exercise above, if we excluded the Lombardy region mortality rate of 1000 per 1 million inhabitants, the Italian mortality rate would decrease to 220 per 1 million inhabitants.

One of the most debated issues in Italy is the number of COVID-19 tests performed [1]; the scientific community is divided by suggestions of a hypothetical high number of undiagnosed cases that would lower the national case fatality rate. This aspect can be easily assessed from the officially declared world data [5]. As reported in Table 2, at the time of writing (April 15, 2020), approximately 1,100,000 COVID-19 tests had been performed in Italy. In the United States, where over 600,000 COVID-19 cases had been confirmed to date, 3,300,000 tests were performed. In Spain (the second most affected country in the world, with over 170,000 total cases), only 600,000 tests were performed; 330,000 tests were performed in France, 1,300,000 in Germany, and 380,000 in the United Kingdom. Therefore, in absolute terms, the nation of Italy has performed the third largest number of tests in the world.

After weighting the number of tests by the population of each nation, Italy has a ratio of 17,800 tests per 1 million inhabitants, ranking first among the G10 countries in terms of tests per 1 million inhabitants (column 3 of Table 2) [5]. This gap is even more evident for the Lombardy region, which has a ratio of 25,000 tests per 1 million inhabitants.

Conversely, if we consider the total of patients who tested positive and the number of swabs performed in the equation, we find that the lower the percentage ratio, the more effectively the nation has tested its citizens.

It is important to observe the temporal misalignment between all the countries because the COVID-19 outbreak started in different countries on different dates. Hence, these values should be normalized by the cumulative number of days since the first outbreak in each country. However, it should be noted that the Lombardy region suddenly discovered the first outbreak with a delay of at least 18 days (ie, 2 weeks of average SARS-CoV-2 incubation time plus 4 or 5 days to develop severe symptoms). In contrast, other European Union countries were able to prepare in advance and may have benefited from a significant advantage in managing the diffusion of COVID-19 by observing the behavior of the COVID-19 spread in Italy, particularly in the Lombardy region.

As already discussed, the Lombardy region has a very significant impact on the national data in Italy. Hence, it may be worthwhile to explore additional considerations to better understand the impact of the Lombardy region on the global data. The Lombardy region has 10,000,000 inhabitants in a territory of 23,800 square kilometers and has a population density of 425 inhabitants/km². Its capital city, Milan, is one of the most dynamic alpha cities in the world, with a population of 1,350,000 people and a population density that reaches 7600 units/km². The Lombardy region also contains the largest logistics hubs in Italy (close to the first outbreak in Codogno [5]) and some of the largest in Europe. It is also one of the regions in Europe with the greatest flow of people and goods by air, rail, or road [10].

In this context, the analysis can be further extended by comparing the Lombardy region data with the most important regions/areas of the other G10 countries (see Table 3). If we focus on the total cases reported, the Lombardy region is the region with the second highest absolute value of COVID-19–positive cases (Lombardy region=59,000; New York City=123,000). If we normalize this value by the total number of inhabitants, New York, Antwerp, and Madrid have the highest numbers of positive cases per 1 million inhabitants. Regarding case fatality rates, the Lombardy region is the region with the highest number of fatalities (18.08%), followed by Île-de-France (17.99%) and Antwerp (14.20%). When we normalize the regional case fatality rate with the national datum (ie, the ratio of the regional case fatality rate to the national case fatality rate), New York City appears to have the highest multiplication factor (9.50%/5.50%) of ×1.73.