Covid-19 deaths: numbers and comparisons between Italy, France, Spain, Germany and the United Kingdom

Covid-19 deaths: numbers and comparisons between Italy, France, Spain, Germany and the United Kingdom

Report Card policy has occurred as the dead are counted by Covid-19 in Italy, France, Spain, Germany and the UK. It is true that each country has its own peculiarities, but it is false to argue that Italy has the least stringent criteria, indeed.

While the second wave of the epidemic is still ongoing, coronavirus deaths have returned to the center of public debate in Italy. If you look at the data in relation to the population, in fact, our country is one of the most affected in the world, with about one death from Covid-19 for every thousand inhabitants.

Several politicians have commented on these numbers, but giving voice to unsubstantiated or very misleading theories, according to which Italian statistics count deaths that would have little or nothing to do with the coronavirus (Sars-CoV-2).

On 8 December, for example, the former Minister of Health Beatrice Lorenzin – now a member of the Democratic Party – said (min. 1:50:14) to DiMartedì on La7 that in Italy "we say that a person died from Covid when has Covid ”, suggesting that the Italian accounts are therefore particularly large. Other states, according to Lorenzin, would take the burden of previous diseases into greater consideration, providing more reliable numbers on mortality.

Official sources in hand, and with the help of our fellow European fact-checkers, we have verified how the deaths from Covid-19 are currently counted in the five major European countries: Italy, France, Spain, Germany and the United Kingdom. And no, it is not true that here we are, so to speak, with a large sleeve in calculating coronavirus deaths, while elsewhere they would be more selective.

Indeed, as we shall see, some of the states taken into consideration even seem to have less stringent criteria than the Italian ones in the calculation of deaths. In general, the statistics on excess mortality in all five major European countries show that deaths related to the health emergency are underestimated, not overestimated.

But let's proceed in order.


Let's start from a premise: attributing a death to single specific causes can often be a difficult process, with inevitable margins of uncertainty. And there is nothing secret in any of this.

As Our world in data also pointed out – one of the most reliable sites to compare coronavirus data in the world, created in collaboration with the University of Oxford – this observation applies to many diseases that can be the cause of death both of patients. healthy than patients with previous pathologies, and certainly not only for Covid-19.

Having said that, how can a “death from Covid-19” be defined as uniformly as possible, to avoid confusion about the data and their interpretations?

On April 20, 2020, the World Health Organization (WHO) published guidelines at an international level on how to define a "death from Covid-19". The criteria are as follows: the death must concern a "probable or confirmed" case of Covid-19, who died of a "clinically compatible disease" with that caused by the coronavirus, unless there are "clear alternative causes of death", like for example a trauma. Furthermore, according to the WHO, there must not be a period of full recovery between the onset of the disease and death.

The guidelines have also been adopted by the European center for disease prevention and control (Ecdc), an independent EU agency that collects data on the epidemic at European level from individual states. But as WHO pointed out in a report of 4 August last, its guidelines are recommendations and each state can in practice follow different criteria to register coronavirus deaths.

Some of these data collection differences have been grouped by the ECDC in a table on its website. A difference in death monitoring, for example, may relate to the criterion of whether or not laboratory confirmation of coronavirus positivity is needed to classify a death as Covid-19. Other countries only consider deaths recorded in hospitals, while others also consider deaths outside the healthcare institutions. Then there are states that set an interval of days from the diagnosis of positivity beyond which a death is no longer counted among those for Covid-19.

But what differences are there, in concrete, in the counts of the five great European countries? Let's start from Italy.



According to former minister Lorenzin, in our country "we say that a person has died of Covid when he has Covid". This is not true, as we have already explained recently .

According to data from the Integrated Surveillance of the Higher Institute of Health (Iss), as of 10 December the deaths of Covid-19 in Italy were over 59,800 . To fit this issue, coronavirus positivity alone – certified with a molecular swab – is not a sufficient condition.

In the Italian definition of “death from Covid-19” three other criteria must be respected , included among those recommended by the WHO: there must be the presence, certified by a doctor, of the typical symptoms of the disease, such as fever and cough; there must be no "clear cause of death other than" Covid-19, such as a road accident or a fall; and there must not be a "full clinical recovery" period between illness and death.

Therefore, the ISS underlined in an in-depth study last November, "the positivity to Sars-CoV-2 is not sufficient to consider the death as due to Covid-19, but the presence of all the conditions is necessary" that we already have listed. To be clear: if a person dies of a heart attack, but in the meantime it turns out that it is an asymptomatic case, positive for coronavirus, it is not counted among the deaths of Covid-19, unlike what Lorenzin implied.

On La7, the former minister Lorenzin brought up the issue of previous pathologies, seeming to suggest that in Italy they are not taken into account in the count of deaths from Covid-19. According to the guidelines followed in our country, the "clear causes" of death other than Covid-19 do not include pre-existing diseases (such as cancer, diabetes or cardiovascular diseases), which may have favored a negative course of the infection. And it is true that almost 66 percent of Covid-19 deaths in Italy have three or more previous pathologies .

But a research published last July by the ISS and Istat – made on a sample of 5,000 medical records – showed that in nine deaths Covid-19 out of ten the disease was the main cause of death. In the remaining 10 per cent, Covid-19 was a cause "which may have contributed to death by accelerating disease processes already underway, aggravating the outcome of pre-existing diseases or limiting the possibility of treatment".

We also remind you that the weekly data of the ISS surveillance system may differ slightly from those published every day by the Civil Protection, since the former, more precise than the latter, may be delayed , according to the communication timelines of the regions ( Figure 1).

Figure 1. Trend of Covid-19 deaths by date of death, updated to December 2 – Source: Iss

In summary: net of some occasional errors, the criteria followed in Italy for counting the deaths of Covid-19 seem to exclude that the official statistics include all the deaths of individuals who have only tested positive for the virus, regardless of other factors.


Beyond these methodological differences – however important – it should be emphasized that the mortality linked to the Covid-19 epidemic is not overestimated, as suggested by those who argue that deaths are inserted in the counts that have nothing to do with the coronavirus. In reverse.

In fact, the data on excess mortality show that compared to the number of expected deaths, those recorded since the beginning of the epidemic in large European countries exceed those contained in the official Covid-19 statistics.

This discrepancy – mostly marked during the first wave – is most likely explained , on the one hand, by the fact that we have missed many coronavirus deaths (e.g. in patients who have not been tested), and on the other hand. side by the fact that the strong pressure on the health system has reduced the ability to cure other diseases, affecting general mortality.

Calculating excess mortality is not an easy task, especially using a uniform methodology across countries. A study published last October in the prestigious scientific journal Nature tried to estimate the impact on mortality in 21 industrialized countries during the first months of the epidemic. In order, England and Wales, Spain, Italy and France all experienced a large excess of mortality.

A similar argument, albeit to a more modest extent, also applies to Germany, not analyzed in the study in question but in other researches.


In recent weeks, we often hear it repeated, even by various politicians, that in the official statistics on the deaths of Covid-19 Italy puts deaths that have little to do with the disease, while in other countries they take into account previous diseases.

We have verified what are the criteria currently in force in the five major European countries – Italy, France, Spain, Germany and the United Kingdom – and the hypothesis just seen is in all likelihood without foundation.

It is true that there are WHO international guidelines on the definition of "death from Covid-19" and that each country adopts different data collection methods. But it is not true that Italy is the one who has the most loose ends in counting, on the contrary.

In our country – and in a very similar way also in Germany – some criteria must be met in order to be counted in the official statistics of deaths from Covid-19. It is not enough to be positive for the coronavirus: you must have a clinical picture compatible with the disease, there must not be a clear cause of death other than Covid-19, and there must not be a cure between the diagnosis of the disease and death.

France seems to be the only one of the five major European countries to count suspected cases among the deaths of Covid-19, not necessarily confirmed with laboratory analyzes. The statistics of the Santé publique, however, mostly report deaths in hospitals and in facilities for the elderly or care, while the analysis of death certificates – more precise and valid for deaths in general – proceeds at a slower pace.

Spain, based on our checks and those of our Spanish fact-checkers, seems to have the broadest criterion for defining death from Covid-19: all confirmed cases of virus positivity are counted among the deaths.

Finally, there is the United Kingdom, which, apart from individual differences between England, Wales, Scotland and Northern Ireland, uses a particular time criterion in the official statistics published by the government. Since last August, the deaths of cases positive to the virus have been counted among the deaths of Covid-19, within 28 days of diagnosis. The national statistical institute has a different collection method, without the time criterion, and also considering suspected cases, where Covid-19 is indicated as a possible cause of death in the death certificates.

Beyond these distinctions, we recall that the data on excess mortality show that the deaths caused by Covid-19 are most likely underestimated, and not overestimated.

(Brief excerpt from an in-depth study published on Pagella Politics; here the full version)

This is a machine translation from Italian language of a post published on Start Magazine at the URL on Sat, 26 Dec 2020 16:18:31 +0000.