The following article is the translation of the Italian ones published on March 15th (http://www.emigrantrailer.com/2020/03/15/corona-virus-covid-19-fattori-di-contagio-e-decessi-in-italia/). Following the latest evolution in Spain, France and Germany, I decided to translate it in English to share this information with more friends and people. The hope is to inform properly about the Italian case, to set up prevention measures abroad.
Just by coincidence, I was in the other part of the world (Singapore and Indonesia) while the situation in Italy and Europe became more critical day by day for what concern the documented cases of COVID-19. It is somehow “funny” to recall that, at the take-off towards Singapore, the information I got was that Singapore belonged to the “risk areas”: 102 cases of COVID-19 (0 deaths) at March 1st (see Table 1). Now, after two weeks, I found myself in a completely opposite situation with several Asian airline companies delivering the information that, starting from tomorrow (15th March), several flights to Europe will be cancelled, and Germany is included: 3062 cases of COVID-19 (6 deaths) (see Table 1). It took two weeks to completely turn the situation around.
In these two weeks (theoretically of vacation and relax as it was planned five months ago), I constantly kept myself updated with friends and relatives in Italy, observing as well the evolution with the frequent news shared by media or social. I have to admit that most of the time I got angry reading news and articles written with low accuracy details, especially reporting numbers and percentages without the proper support of information. In these two weeks, I promised myself to avoid any post or article on this topic, since there was already enough chaos in the media.
However, after reading some contradictory information, I decided to execute my personal analysis, driven by methodologies developed in 15 years of scientific research and work. Of course, I started from the official data sources (World Health Organization and/or Italian Government) and I spent part of the time of my vacation to study and analyze all these data.
Let´s start from the Italian situation (see Table 2, where the cells highlighted in orange indicate the values above the national average).
The data updated at March 13th clearly indicate a situation with the highest criticality in the regions of Lombardy, Emilia Romagna, Veneto, Piedmont and Marche. Like many Italians, I asked myself why the evolution of cases of COVID-19 in Italy has been faster in these regions. To find the answer, in addition to the official data by the Italian government (see left columns in Table 2), I have added three important factors: the population, the population density and the number of municipalities. These are important parameters to convey the message of latest weeks: stay at home and limit travels as much as possible.
The reported values indicate that Lombardy has basically double the population of the other regions, the second highest population density in Italy (but it would be the highest without the low contribution of the district of Sondrio), and has the highest number of municipalities. In other words, the contact among people is statistically more likely. Specifically, the two provinces of Monza Brianza and Milan are at the second and third place for population density in Italy, respectively with 2156 inhabitants/km2 and 2063 inhabitants/km2. To compare, the population density of the Chinese province of Hubei, where this virus originated, is 308 inhabitants/km2 (see Table 3). We are talking about 7:1 ratio. Namely the two Lombard provinces have the population density seven times higher than the one of the Chinese provinces. Bergamo (405 inhabitants/km2), Varese (743 inhabitants/km2) and Como (463 inhabitants/km2) are other three Lombard province in the Italian top ten for population density and their values are still above the one of the Chinese provinces. If the message is not yet enough clear, let´s focus for a moment on the major cities of these regions: Milan (Lombardy) has a population density of 7550 inhabitants/km2, while Wuhan (Hubei) has 1200 inhabitants/km2. Again, a ratio 7:1. Maybe, it is clearer now why the contagion´s spread in Lombardy has the highest rate of growth.
In these days, in Italy, several opinions indicated that the highest number of cases is linked to the highest number of medical swabs. This is true, but only partially. Namely, if a direct correlation exists, the Emilia Romagna would not be in the second place for cases with COVID-19, showing a significantly higher number of cases compared to Veneto (see Table 2). Simultaneously, how could be explained that Piedmont have three times the number of cases of Lazio, having half of the medical swabs performed? In other words, the number of swabs is important, but not decisive to understand the special situation in these northern regions of Italy.
Instead, another factor should be considered and it differentiates Lombardy from the other regions (see Table 4). In fact, this region has a number of daily passages of people (airports, trains, roads, highways, etc.) not comparable with all other Italian regions.
For example, as shown in the Table 4, in the last January, the number of passengers in the Lombard airports was the highest. A fair objection would be: but the number in the Lazio airports is not so different! That´s true. But it is also true that this type of traffic is localized in the Rome Fiumicino Airport (FCO) with 2742700 passengers, since the contribution of Rome Ciampino (CIA) is marginal with 467467 passengers. Instead, in Lombardy, this traffic is spread among three major areas: Milan Malpensa (MPX, with 1918629 passengers), Milan Linate (LIN, with 643430 passengers) and Bergamo (BGY, with 1053599 passengers).
Moreover, Lombardy has 115 km of metropolitan network versus the 59 km of Lazio and 30 km of Campania. We are talking about twice the values, with a greater efficiency. Following this factor, it is noteworthy to remark that Lombardy is at the center of a complex and numerous motorway and railway network, which connects both northern and southern parts of Italy, as well as western and eastern parts of Italy. If we observe the numbers of vehicles and the number of kilometers of the railway network belonging to these four neighboring regions (with Lombardy in the middle), it is evident how higher the values are compared to the mean national value.
In conclusion, it cannot be a surprise how fast the contagion´s spread is in these areas if we sum both reported factors: population/population density & connectivity/networking of the transport infostructure within and outside this region. As consequence, the other neighboring regions are also affected like Marche, that in theory would have these factors within the mean national value. In fact, it is not a coincidence that about 70% of the cases of COVID-19 documented in this region belong to the province of Pesaro and Urbino, at the border with the Emilia-Romagna.
These data should help to understand how important and necessary is to limit contacts and to support isolation/quarantine as much as possible, especially for those who live within these regions. This message is not clear at all to all those persons that in the last hours are “assaulting” the trains to move from northern regions to southern regions, becoming a key factor to spread the virus in areas where the situation is not yet so critical. In fact, for instance, if the situation in Campania & Lazio is still somehow under control, it depends mainly on the fact that the daily connectivity/networking is not so high as in Lombardy. This factor slowed down the contagion´s spread in these two regions, where other factors (population and population density) are similar to those in the norther regions, even if more centralized in areas like the city of Rome and Naples.
Once completed the analysis of the contagion´s factors, I tried to give an answer to this second question: why does Italy have a percentage of deaths higher than in other nations?
In these days, I read several opinions. Some shareable, some rather generalist (without support of any data). Other completely nonsense.
Many underlined that there are several ambiguities in the comparison of different nations data: the methodologies to count deaths (some countries count only if the COVID-19 is the unique responsible, other even if the COVID-19 is contributing to other factors), the number of medical swaps or the fact that countries do not officially declare cases and deaths. Leaving behind any speculation or conspiracy theory, the truth is that a scientific approach would require a detailed analysis of each single case, in order to quantify how much the virus has contributed to the death of a person.
In this regard, I wish to underline that the root cause analysis for any death is under the responsibility of medical doctors or medical specialists, who have the required scientific background and experience. I limit myself to analyzing the published official numbers.
Personally, I think that a great mistake reproduced several times by the media in the last two weeks is to keep looking at percentage values, without having a clear, established and available denominator (the real number of infected persons). This approach is wrong when comparing the status of the different Italian regions. This approach is wrong when comparing the Italian status with the other nations.
Let me explain better using the data in Table 5.
Up to now (in the moment I write), the official number of deaths in Lombardy is 890. This is the only official and stable data that everybody knows. Now, let´s start to generate percentages, with a focus on this region: if we calculate the percentage of deaths versus the number of cases, we will get 9.1 % as value; if we calculate the percentage of deaths versus the number of medical swabs, we will get 2.7 % as value; if we calculate the percentage of deaths versus the total population of the region, we will get 0.009% as value. I could continue forever generating thousands of different percentages, simply changing the denominator.
Every percentage value would result different from the other and could be used according to the scope of the message to be communicated. In other words, the percentage value can give an idea but it must be always supported by the total number. To be more explicit, in the very first period of the COVID-19 crisis in Europe, between 16 and 26 February 2020, the number of COVID-19 cases in France was 12, with 1 death. If we calculate the percentage value, we obtain 8.3%! This number is not far away from the current percentage values in Lombardy and Emilia Romagna. But obviously no one at that time got afraid, because it was “only” one death (“only” written like this because every death, even just one, is important).
Therefore, journalists, commentators, politicians, simple citizens, should start from the fact that nobody has the knowledge of the real number of infected persons, and whatever percentage we calculate versus other denominators, it should be treated and communicated carefully. Possibly, we should always accompany it with the total number, published or communicated by the official responsible departments or agencies. Furthermore, there are still a lot of cases under hospitalization or intensive care and these numbers could improve or get worse on the specific evolution of these cases.
Having made this necessary premise, I return to the initial question: why does Italy have a percentage of deaths higher than other nations?
In these days, I read a lot about the fact that Italy is a “old” nation, and that the elderly people (over 65 years) is the category most at risk.
In all this context, I was not able to see any data supporting this statement. Thus, I decided again to do my own research and the data are shown in Table 5. Using as data source the official ISTAT data of 2019, Lombardy has about 2 million of over 65 residents, a number much higher than other regions and representing the 16.5% of the entire population of over 65 people in Italy.
Extending the analysis, the other nations, it catches the eye the big difference compared to China, Republic of Kore, Iran and USA (United States of America), all of them much “younger” than Italy and Lombardy (see Table 6).
Waiting for the final scientific confirmation (but official initial report from ISS – Istituto Superiore di Sanita´ indicate that most of deaths are elderly people with about 80 years old and already pre-existing health issues), the age seems to be a key element to understand the different rate of deaths among the nations and the regions.
It could be a fair objection saying that France, Germany and Spain have a percentage of over 65 not so far from the Italian ones, but these nations do not show the same rate of deaths. To explain this aspect, it would be needed a deeper analysis of the other three nations, as executed for the Italian case. I would not be surprised (but I hope to be wrong) that in the next days these nations will experience a similar situation, especially if a significant number of infected persons reaches highly populated areas like Paris in France (ca 21000 inhabitants/km2), Madrid (ca 5300 inhabitants/km2) and Barcelona (ca 16000 inhabitants/km2). It could be different the German case, where the biggest metropolitan city is Berlin, with a population density of about 4000 inhabitants/km2, but without a similar connectivity/networking transportation system like the others. However, a dedicated analysis is still on-going.
Summarizing, in this new scenario that Italy and Europe are nowadays facing, the important factors to consider simultaneously in the context of contagion´s spread and deaths by COVID-19, are:
- The most populated areas
- The areas with the busiest, largest and most efficient transport systems (train, roads and airports), both for internal (national and regional) and external (international) travels
- The areas with highest number of elderly people (over 65)
Therefore, it is very important to respect the isolation or quarantine protocols in order to reduce the spread as much as possible in more populated areas.