Talk:COVID-19 pandemic in Italy/Archive 3

Foreign cases linked to Italy
Bulgarian here. There is no proof that first Bulgarian cases are linked to Italy, although it's believed so in my country due to intensive transport connections between both countries. Also I couldn't find any reference from the health authorities, media or other reliable sources confirming such a link. I think the Bulgarian cases should be removed from the Foreign cases linked to Italy. 92.247.60.126 (talk) 21:11, 19 April 2020 (UTC)
 * Done. Toddy1 (talk) 23:06, 19 April 2020 (UTC)

Also, the discussion on this section was obviously never concluded; it was pushed around to the project talk page and then archived away. And we are still left with that monstrous section, which frankly, serves no actual purpose. This thing could be paraphrased as "there were cases in x countries linked to the Italian cluster", and not lose anything relevant. This is an encyclopaedia, not an epidemiological logbook. But somehow someone is apparently attached to this; and I wonder how we can move forward. Averell (talk) 20:37, 26 April 2020 (UTC)
 * The discussion was never concluded because there were no concrete solutions. The section is so big, because Italy was as far as we knew at the time the main spreader in Europe and around the world. The sources are there. My suggestion to move it into a subarticle was not taken up, but that would be the only solution to reduce the section here without loosing content. The same type of section is in many other country articles as well, just a lot smaller, as they did not have the impact that Italy had. Agathoclea (talk) 20:57, 26 April 2020 (UTC)
 * There are four solutions: delete the section, since with two more editors (you guys) saying it is worthless then consensus has been achieved, drastically shrink the section to a few examples and the map, convert the list structure into prose (which should dramatically shorten it), or spin it off into its own article and allow it to fester and die on its own. If the section owner reverts, he will be outnumbered (at least) three to one and cannot successfully revert without edit warring. Abductive  (reasoning) 06:26, 27 April 2020 (UTC)
 * I have not said it is worthless. Actually the opposite. In fact so far there has been a consensus to keep the material. It was only the question of presentation left. Please do not start your editwaring again. Agathoclea (talk) 06:37, 27 April 2020 (UTC)
 * At the very least, the consensus was to prosify the material. The way it is now, it will keep attracting negative attention over time, as the post by Averell23 shows. Why not do what you want, and move it to a subarticle? I will support that, but you have to do it because the section owner will revert me. Abductive  (reasoning) 07:10, 27 April 2020 (UTC)
 * In addition, most of the sources are questionable... the sources should be epidemiological studies and not newspaper articles where it is said Y was at X, so he took the virus at X, without even a genetic analysis to understand which strain of the virus it was and if and where the same strain was circulating. And this apply not only to this specific section but also to similar section like the Spanish one. Also talking about Spain, a study found 15 strains of the virus circulating in the country which imply multiple introductions of the virus and not a single patient 0... the situation is probably the same for most of the countries, so citing the spread form one country to another in relation to a single person is misleading in my opinion, and gives a distort view of how the virus actually spread.--80.182.168.40 (talk) 07:54, 27 April 2020 (UTC)


 * again please stop with this language, there is no section owner: the history shows that there was one editor – you – who was edit warring to remove an entire well-sourced (secondary-sourced) section multiple times, and a few editors successfully opposing them, as anyone can read in the archive of this talk page. None of your proposals has reached consensus, not even the one to make it prose: the discussion simply died because nobody was so convinced about the need of transforming a regular list (Wikipedia is full of lists) into "prose". However as I said, I am not against the idea – I simply wouldn't do it myself.
 * This is an encyclopaedia, not an epidemiological logbook. This encyclopedia is literally full of COVID-19 timelines of all sorts, Timeline of the 2020 coronavirus pandemic in the United States, or Timeline of the 2020 coronavirus pandemic in Spain for example, with day-by-day lists of what happened each day, with a level of detail such as "3 March Another infected person was reported in Málaga, Andalusia, with a total of 13 reported cases.[66]" (from the Spain timeline). I don't see the difference between such a time-ordered list and a country-ordered list where again, all that is reported is well-sourced, and – in order to appease Abductive's wishes – there are even citations at the top of the section where it is generally stated (with a few examples) that the virus was indeed spread from Italy abroad.
 * International news organizations like Reuters are perfectly reliable sources when stating something like "the case XY came from Italy". The section is not about the genome analysis of the virus, it's about, as the title says, "foreign cases linked to Italy". --Ritchie92 (talk) 08:09, 27 April 2020 (UTC)
 * In fact, I am questioning the existence of this section (and other similar ones in other pages), because in the midst of a pandemic knowing, for example, that 10 of the 80,000 Chinese cases are related to Italy (or that 3 of the 24.000 Portugal cases are related to Spain) does not add anything to this pages (if not weight) while this is an information that should be contained in the other countries respective pages (which for the most part is already the case).--80.182.168.40 (talk) 11:20, 27 April 2020 (UTC)
 * The moment somebody other than me removes this garbage, it will be the end of it. Abductive  (reasoning) 18:03, 27 April 2020 (UTC)
 * Content that is well-sourced and with multiple citations is not "garbage". It is a perfectly acceptable content on this encyclopedia, it's on topic, it's relevant, it's something that has been treated in multiple articles by many international and well-respected news agencies. I don't see the problem in having that section. --Ritchie92 (talk) 18:46, 27 April 2020 (UTC)

Yes - this encyclopedia is literally full of COVID-19 timelines of all sorts. And yes, those other parts are also a mess. Most of the COVID articles are a mess to some degree. That doesn't make this mess any better. Frankly, there seems to be an understanding that "reliable sources" is the only threshold for inclusion, but what information is this section supposed to convey? Just look at the section about "France": "A man was infected, and another man, and his daughter..." and so on. With 15 (!) inline citations for that one paragraph. But does it add anything to my understanding of COVID-19 Pandemic in Italy? So, my question again, how should we proceed? Abductive has actually proposed several possible solutions, of which deletion was only one, and each would be a vast improvement. Nobody has indicated any preference or any other compromise that isn't "keep it as it stands". Averell (talk) 19:54, 27 April 2020 (UTC)
 * Exactly. There are many easy solutions, and one slightly harder one of rewriting the list into prose. Whoever cares most about the material surviving needs to make a decision and do the work, because over time, more and more editors will decide deletion is the optimal solution.  Abductive  (reasoning) 20:06, 27 April 2020 (UTC)


 * Sorry but this is wrong: I wrote already multiple times that switching the section to prose-style is fine for me, and there was actually another editor who started doing some work in this sense. Wikipedia is a work in progress and there is no rush to implement such a stylistic change immediately (with the threat that other editors could otherwise delete the section).
 * Regarding your comments, the fact that a topic is treated in many news outlets makes it definitely notable and therefore per WP standards, worthy of mention in a WP article. The topic is clearly linked to the 2020 coronavirus pandemic in Italy, and therefore here it is, appearing in the related article on WP. I don't see how can you argue the uselessness or the irrelevance of the information that in some cases – at the very start of the pandemic at a world-wide stage – there was spread from one country to many others. I think it is a crucial information regarding the spread of the virus worldwide, it was treated by the press in many countries, and it was on the first pages of the main news network of Europe for days. So the section should definitely be kept, and if there are some parts (like the French one that you mentioned) that are weird or simply "too much" let's reduce them. But a clean cut of the entire section as a solution to the section's problems is not the way to proceed. The section needs improvement, but is not irrelevant to the article's subject.
 * Currently even the banner that Abductive inserted on top of the section is plain wrong. All sources cited in the section are secondary sources. --Ritchie92 (talk) 20:12, 27 April 2020 (UTC)
 * Okay, if prose-style is fine for you, what would you like to keep of the section? More to the point, what is the topic of the section? If the topic is "there were cases in many countries linked to Italy": Yes, that should be included in the article. It can be done in one paragraph (IMHO), and certainly without listing every case, their nationality, grandmother and which hospital they were treated in. Averell (talk) 05:54, 28 April 2020 (UTC)
 * I agree in reducing the size, probably one paragraph per each continent with some detail on the countries where there were cases. And I agree in leaving the grandmothers out of it. This was started to be done by, I don't know if they are willing to continue, otherwise when I have time I could do it. It won't take me 10 minutes, to be clear. --Ritchie92 (talk) 08:22, 28 April 2020 (UTC)
 * The main issue is that not all cases are the same, else we could say x,y, and z were seeded from Italy. In some instances it was the first or one of the first known cases. In some instances it was a major contributor to new clusters. In other instances it was identified on arrival and isolated. These are the types of information that need to be retained and i find them easier to identify in bullet points rather than a long paragraph of text. I still think with the volume of information a spinout article might be worthwhile but I cannot come up with a good name for it. Agathoclea (talk) 08:44, 28 April 2020 (UTC)
 * I don't think the volume of information is so large to require a split to another article. After all, probably a lot of information that is currently in the list should go out (as Averell was suggesting), so it won't be that large after a clean-up. And I agree that a title for such an article would be extremely specific and weird; I would bet that in a few weeks someone will notice the abnormal specificity of the new article, there will be a merge request and it will be merged back into this article. --Ritchie92 (talk) 08:54, 28 April 2020 (UTC)
 * Another possibility is instead of trying to preserve continents, which are irrelevant to a disease vectored by airplane travel, you try to preserve the timeline. Abductive  (reasoning) 23:55, 28 April 2020 (UTC)

First off, converting into prose and reducing the size sounds great. As for the Agathoclea's argument: I don't think that the "type" of case should be very relevant here. The only thing common to these cases are that they are related to Italy. While there is apparently consensus to keep the information, we need to be careful not to present it in a way that would imply epidemiological conclusions (such as "the country was seeded from Italy" or "it spread to n countries from Italy") - these kind of statements aren't actually supported by the sources and would land us in WP:OR. Also, if one of the cases is in fact crucial to that country's outbreak, the details should rather be in that country's article rather than here. Preserving the timeline would work well for me, as it'd be both natural and neutral. — Preceding unsigned comment added by Averell23 (talk • contribs)

Meaning of the statistics
There is a dispute between editors: Ritchie92 prefers the simple version.  Oll.21 prefers the complex version. Perhaps this should be discussed here. Toddy1 (talk) 13:50, 19 April 2020 (UTC)

Regarding the "Statistics" paragraph in the 2020 coronavirus pandemic in Italy, the information present before my edit was biased and drew wrong conclusions from the WSJ article cited (the only source there was). The article is biased, coarse and sensetionalist: the fact that two cities in Lombardy may have underestimated the number of deaths from coronavirus does not mean that the stats for Italy in general are "considered underestimates". Moreover, the first sentence does not have a source, probably because it is wrong. In fact, our death stats do include people who died outside of hospitals, even though, of course, many are missed. I modified my original edit, and removed my own analysis of the situation, to avoid violating the no original research policy. What is left is factual, indeed the notion that death toll has been severely underestimated in many European countries is well established and reported by Italian news outlets. You reverted my edit twice, accusing me of violating the no original research policy. I referenced every sentence I made, without including my own analysis. If you keep reverting my edits, putting your own opinions/views before the evidence I brought, you will be reported to an administrator. — Preceding unsigned comment added by Oll.21 (talk • contribs) 12:41, 18 April 2020 (UTC)
 * Copied from User talk:Ritchie92


 * The fact that the WSJ article is biased is your personal opinion. Your edits instead, apart from being a form of edit warring, draw a conclusion which is the opposite than the one coming from a reliable (and third-party) source like the WSJ. You use a far-right nationalist newspaper as Il Primato Nazionale and the website today.it as source, which are far from being reliable (or as reliable as the WSJ), and – concerning the first one – of course having a partisan view. For such a delicate topic as a worldwide pandemic, I suggest using reliable sources, in this case I would rather use sources that are not Italian newspapers, which might have a second political intent to push for a certain view on the subject. Finally, I suggest you use the talk page of the article to discuss, instead of my talk page. --Ritchie92 (talk) 17:43, 18 April 2020 (UTC)


 * There was an article published by Hromadske on 23 March that might be relevant. This said that Italian statistics recorded coronavirus as the cause of death, when a patent died if he/she was confirmed to have the coronavirus, but that inspections of the Higher Institute of Health have showed that only in a minority of these cases was the virus the direct cause of death.   Toddy1 (talk) 14:45, 19 April 2020 (UTC)


 * Unfortunately I cannot read Ukrainian. However, I would say there there are two contrasting issues. (1) The number of deaths of people with coronavirus is underestimated in Italy, as the WSJ says, as it probably is in many other countries, due to the limited amount of testing available, and to the number of non-tracked deaths in nursing homes, at home, etc. again as the article reports; this level of underestimation is comparable to the comprehensible underestimation of the total number of infected, where the main issue is related to the testing capabilities of a country. (2) It is also true that Italy counts into the amount of deaths all people who die with coronavirus (so, not only because of it) – obviously except the ones that go under the radar of the limited tests performed; in this case this is unlike some other countries, which adopted other criteria as counting only deaths in hospitals, or only deaths because of the virus, etc.
 * However, in any case, this does not put Italy to a higher/better position with respect to the counts that are done by other countries, as the sentences introduced by the other editor seemed to suggest. I think both things that I mentioned here can definitely be added to the text in a couple of sentences, and in a properly-sourced way (e.g. using international reliable – possibly English-language – newspapers, magazines, or journal articles), but the version pushed by the editor was poorly sourced, drawing personal conclusions, and contradictory (and I am not focusing on the edit-warry behaviour). And finally, I will never accept the use of a far-right white nationalist newspaper as Il Primato Nazionale as "reliable source" in an article related to a medical subject. --Ritchie92 (talk) 17:29, 19 April 2020 (UTC)
 * In March 2020, Walter Ricciardi (who is an advisor to the Italian Minister of Health) said "The statistics are regional and very often indicate coronavirus as the direct cause of death. However, as inspections of the Higher Institute of Health showed, only in a minority of cases did the virus directly cause death, although it undoubtedly contributed to the deterioration of the patient." Silvio Bruzaferro (president of the Higher Institute of Health), said that 48.5% of those who died from [with?] coronavirus had three or more comorbidities; 25% each had two or one; less than 1% of those who died did not have any other chronic illness.


 * The meaning of the statistics from any country is a problem. If I understand what has been said about the English data correctly, it has timelags.  So when the number goes up by 900, we do not know how many of that 900 died in the previous 24 hours.  Many people had been assuming that it was 900, but some of them might have died a week earlier but only joined the daily total once it had been confirmed that they had the disease.  Toddy1 (talk) 18:10, 19 April 2020 (UTC)


 * Many thanks Toddy1 for starting this discussion, I am new to this side of Wikipedia (the editing side) and I appreciate your mediation. There are two reasons why I felt the need to make the edit. 1) The first sentence of the previous version contains false information: "Deaths statistics for Italy do not include the total amount of coronavirus victims who died outside hospitals". This is simply incorrect. Post-mortem tests are carried out and people who died outside of hospital are indeed included in the death toll. Of course, not all of them are, and it is much more likely for a person dying in hospital to be included in the death count than it is for someone dying outside of hospital, but this is the same everywhere. Other European countries do not hide the fact that they are only counting hospital deaths (e.g. UK and the Netherlands, hence why I mentioned this in my version). 2) The second sentence is misleading and based on an overly sensationalised article trying to sell the results of an analysis on a couple of cities as a general problem in Italy's death tally, as evidenced by the title, which is nothing short of cheap clickbait. Ritchie92 seems to think that the article is reliable based entirely on where it's published. I respect your opinion, but I have actually read the article, and can say that the "analysis" they make concerns three cities in Italy's worst-hit region (Coccaglio, Bergamo and Brescia). Now, there is quite a jump between saying something like: the death toll is likely underestimated in the worst-hit areas of the country, to saying that "the official deaths statistics are considered an underestimate" for the whole of Italy. I would like to remind you that the healthcare system became overwhelmed in Lombardy, there is no other region where this has been the case.
 * On a more general note, following on from Ritchie92's reply to my message, I wonder why they feel uncomfortable using Italian newspapers as sources, on a Wikipedia entry about the coronavirus pandemic in Italy. Seems a bit counter-intuitive to me. Again, I respect your opinion when you say: "In this case I would rather use sources that are not Italian newspapers", but that's just what it is, your opinion. I would rather use them, since it is obviously more likely for Italian newspaper to report on this than foreign ones. If you can offer concrete evidence to show that Italian newspapers, in your words, "might have a second political intent to push for a certain view on the subject", then I will be happy to reconsider my entry. Until it is just your opinion/generalisation, please do not put it on a higher level than the multiple articles I cited, from a variety of newspapers.
 * Lastly, I would like to respond to Ritchie92's interpretation of my entry. In one of your replies you wrote: "this does not put Italy to a higher/better position with respect to the counts that are done by other countries, as the sentences introduced by the other editor seemed to suggest". In my original entry, I accept that this was indeed the message that came across, contrary to my intention. In the updated version that you read above, however, there is no implication that Italy's death count is better than that of other countries, I simply stated that it is not comparable. I will also say that, in the case of countries who openly just report hospital deaths, it is clear that the Italian death count is more accurate. Nothing wrong with stating that. In your last reply, you seem to brush off the fact that some other countries count deaths differently: "in this case this is unlike some other countries, which adopted other criteria as counting only deaths in hospitals, or only deaths because of the virus, etc.". Well, this is quite a big deal, actually. Over 96 % of deaths in Italy had one or more pre-existing conditions, so forgive me, but it is just not good enough to brush off the fact that some European countries do not include many of these deaths, because they may not have been caused by coronavirus. Again, there is nothing wrong with pointing out this issues. I read countless articles, both in the Italian and foreign press, making all sorts of comparisons, taking the numbers as they are, without providing any context or critical evaluation of them. For example, UK-based newspapers casually comparing the UK death toll to the Italian one... Simply ridiculous (see above)! I find this quite frustrating, and I believe it is right to briefly point to this issue in the section I edited.
 * I am happy to concede that including "Il Primato Nazionale" in my references was not a good idea, and would be happy to remove it. But the general message of my entry stands. — Preceding unsigned comment added by Oll.21 (talk • contribs) 20:00, 19 April 2020 (UTC)


 * edit conflict Citation 3 Il Primato Nazionale (10 March 2020) does not seem to mention any of the stuff it is cited for. What it does say is that Carlo Fidanza (a politician) said: "in some countries, such as Spain, post-mortem swabs have shown that some people classified as victims of normal flu actually died from the virus."  Apparently Fidanza was suspicious that the ratio of dead to infected in Germany was too low to be true.  I realise that Ritchie92 does not think that Il Primato Nazionale is a reliable source.  Toddy1 (talk) 20:03, 19 April 2020 (UTC)


 * Citation 2 Europa Today (30 March 2020) more or less supports what it is cited for. Toddy1 (talk) 20:16, 19 April 2020 (UTC)

I understand that you are new to editing on Wikipedia, however there are some rules to follow. First of all, do not engage in edit warring when one of your edits is reverted: this means that the edit is controversial and you should proceed following the so-called "bold-revert-discuss" cycle. Second, I would suggest you not to try to give a personal interpretation of the sources and/or synthesize material taken from different sources. In general, don't edit on Wikipedia just because you feel personally frustrated about something or want to right great wrongs: it is likely that your edit could be perceived as original research and tendentious.

Coming to the issue, I want to stress again the importance of reliable sources. The Wall Street Journal is a perfect example of a reliable source as of Wikipedia standards; so again calling the WSJ article "nothing short of cheap clickbait" is your personal and probably very biased opinion. On the contrary, Il Primato Nazionale is the newspaper linked to the Italian neo-fascist organization CasaPound. This means it is definitely not a reliable source, we don't need fascist propaganda on Wikipedia. I wonder why they feel uncomfortable using Italian newspapers as sources I don't feel uncomfortable using Italian sources in general. However there are two reasons why I would prefer using English-language sources (but it's not mandatory): (1) this is English Wikipedia, so a non-English source should be used as a last resort, only if there is no equivalent and reliable English-language source; (2) in this case the topic has become political in Italy and we simply cannot fully trust all Italian news organizations (especially the ones chosen by you that are a fascist newspaper and a not very prestigious web news source today.it; I would suggest using actually good agencies like ANSA.it or serious newspapers like Il Corriere della Sera or Il Sole 24 Ore): therefore I think it would be probably much easier to find more neutral point of views on the best foreign press.

Regarding the content issue, I support having a few more sentences describing the comparison with other countries, however we must still stress the main general point: i.e. that the death count is an underestimation (if it's underestimated in a few cities, yes, it's underestimated overall). This is confirmed not only by the WSJ, but also by Bloomberg, Reuters, the Italian newspaper Il Corriere della Sera, and the Italian National Institute of Statistics ISTAT, and I only cited the most relevant sources. --Ritchie92 (talk) 09:04, 20 April 2020 (UTC)
 * Thanks. This is useful. Please could you point me to the page on the ISTAT site that says about underestimation of the number of deaths. The link you gave was to a more general page.  The page on Characteristics of patients who died positive for SARS-CoV-2 infection in Italy is super. -- Toddy1 (talk) 10:48, 20 April 2020 (UTC)
 * ISTAT does not directly say that, as the national statistics agency it just provides raw data, however you can see an analysis of the raw ISTAT data here, where it's clear that the number of deaths is far higher than reported. --Ritchie92 (talk) 08:40, 23 April 2020 (UTC)

Thanks for your replies. Ritchie92, I will not respond to your "right great wrongs" comment, as if you had read the Wikipedia entry you yourself linked, you would know that it isn't relevant (none of what I wrote is my "original thought", as evidenced by the references I used). Instead, I will dive right into the content issue. As I said, I acknowledge that one particular source I used ("Il Primato Nazionale") is not appropriate, and I am ok to remove it. I propose an updated "working version" which you can see in the table above. I rearranged my original edit, and nuanced most sentences. I also added a few more references to both Italian and foreign news sources. As I said before, I do not agree with your preference not to use Italian sources, so I kept most of them. As for the "main general point", I don't disagree with you on the underestimation of the death toll. What I do not agree with is the notion that if the death toll is underrepresented in Lombardy, it must equally be underrepresented in all other regions. Once again, the healthcare system was not overwhelmed anywhere else, so it is likely that the underestimation (if any) was much less severe in other regions. Every source you referenced in your reply, including a random tweet (bit of an odd choice there, imho), are based on analyses including the most badly affected towns, mostly in Lombardy. Indeed, one of the most comprehensive analyses comparing death tolls in different countries, published in The Economist, only includes data for Lombardy. The same tweet you referenced cautions: "Ci sono 1689 comuni (non rappresentativi dell'Italia)" (translated: "There are [the stats include] 1689 municipalities (not representative of Italy)"). I hope the new "working version" tackles some of the issues that have been raised, I am of course happy to discuss any edits and/or additions to it. — Preceding unsigned comment added by Oll.21 (talk • contribs) 13:12, 23 April 2020 (UTC)


 * I do not agree with your preference not to use Italian sources, so I kept most of them, fine but if there is no equivalent English-language source the Italian ones should be removed, as per WP:NONENG.
 * What I do not agree with is the notion that if the death toll is underrepresented in Lombardy, it must equally be underrepresented in all other regions. I will explain you with an example. Let's say Lombardy has 100 real deaths, of which 60 are officially counted, and the rest of Italy also has 100 deaths, all officially counted. Then the official count for Italy would be 160, while in reality there are 200 deaths, and 160 < 200. So obviously if there is an underestimation in a region of Italy, it means that there is an underestimation also in the total national count. This is basic math, so I don't understand what are you disagreeing with. A sentence stating that "the total death count in Italy is underestimated" is technically true (and well sourced).
 * including a random tweet (bit of an odd choice there, imho) I just used the tweet after having a request for a synthesis of the ISTAT data from another editor. I was not and I am not suggesting to use that tweet as a reliable source in the article text.
 * I have done a few corrections to the working version, however I might work more on it later (and add more sources). In particular I removed the unsourced sentence: "At present, it is unclear whether this has been the case in other Italian regions, and if so, the magnitude of the death count underestimation is unknown." which is an editorialization and an undue WP:SYNTH of what is stated elsewhere. I also removed the final sentence: "For these reasons, death tolls are not comparable among countries." which is also an editorial comment (and synthesis), and also unsourced. I still do not understand why there is the need of having a description of how the method of counting works in other countries, whereas this article is about the pandemic in Italy. It looks to me like the editors are pushing for transmitting the non-neutral POV conclusion of "Italy is counting deaths better than other countries" which is definitely debatable, and definitely at the moment not the kind of information that should go on Wikipedia, since it is not proven nor stated by authoritative sources (for example, no one at the WHO has reprimanded Germany or France or Italy for how they count the deaths). So I would remove or drastically reduce the parts where it is described what other countries do. --Ritchie92 (talk) 18:46, 23 April 2020 (UTC)


 * Thanks for your response, Ritchie92. I understand what you mean about the total number of deaths being an underestimate, and I am not opposed to adding a sentence stating that. However, I think another sentence should follow, stating that the analyses showing this underestimation were done on a subset of towns, mostly in the worst-affected region of Lombardy, and therefore it is not known if the underestimation of the Italian death count is of the same magnitude as that suggested by said analyses. This is an important point to make. You are right to say that if one region underestimates the death toll then the total figure for Italy is an underestimation, but I highly doubt that the total death count is underestimated by ~ 50 % (as has been suggested for Lombardy).
 * I am fine with the removal of the phrase about death tolls not being directly comparable among countries, even though I still personally believe there is nothing wrong with stating the obvious. I think the edits you made are fine, and I am happy with this current version.
 * The conclusion you make from my proposed edit (that "Italy is counting deaths better than other countries") is quite a far-fetched interpretation, imho. Fair enough, you may have got that message reading my first version, but this revised version does not suggest that at all. It states the obvious, i.e. that death tolls are not directly comparable. Something that has not been pointed out enough. Yes, this is the Wikipedia page for the coronavirus pandemic in Italy, but context is important. I even cited a source suggesting that Lombardy is actually one of the worst regions analysed in terms of % of excess deaths captured in the covid-19 statistics. So again, I don't see how you can conclude that from reading this revised version. --Oll.21 (talk) 18:48, 27 April 2020 (UTC)


 * Ok, I just added another small sentence that briefly explains that the estimate is made in some areas of northern Italy and with the excess deaths with respect to the seasonal averages of the previous years. If you are fine with it I will then put it in the article as it is now. --Ritchie92 (talk) 09:08, 28 April 2020 (UTC)
 * Thank you Ritchie92, that reads well, I am happy with it. --Oll.21 (talk) 16:23, 30 April 2020 (UTC)

Edit request: Include graph of daily mortality from all causes


The national institute of statistics (ISTAT) publishes and updates weekly a dataset of daily deaths from all causes for the years 2015-2020:

https://www.istat.it/it/archivio/240401.

From this data it is possible to plot daily deaths by year: [edit: graph linked below is obsolete, see figures on this page]

https://raw.githubusercontent.com/allaisandrea/mortalita-covid/master/figure.png

The results are striking: there is an unprecedented peak for this year (2020) starting at the beginning of March. Given the prevalence of the opinion that COVID-19 is "just like the flu", I think it would be useful to include this plot or a similar one in the page, to emphasize the exceptionality of the COVID-19 outbreak. However, I cannot publish the graph myself because the page is protected and my account is not confirmed, so I am appealing to the administrators.

Allais.andrea (talk) 21:27, 4 April 2020 (UTC)
 * You need to define "all cause" in this graph, because seems like this graph suggest the 60 millions Italians alive at the moment have a life expectancy of about 400 years. Very interesting data, if we substract COVID from this one, this suggest an abnormally low case of other infections, which is a great news. Iluvalar (talk) 23:53, 4 April 2020 (UTC)


 * Good observation Iluvatar. The explanation is that the dataset does not cover the entire national population, but only a subset of about 1000 cities for which data was available at the time of publication. With some work it would be possible to figure out what fraction of the population those cities represent, but I have not done it yet. Allais.andrea (talk) 18:43, 5 April 2020 (UTC)
 * Allais.andrea, the figure is very interesting but seems to disagree with the z-score graph for Italy from https://www.euromomo.eu/, for year 2017. Why would that be? --Dan Polansky (talk) 16:03, 7 April 2020 (UTC)
 * More in the line of the Iluvatar comment: per ourworldindata.org, in 2017 Italy had 608455 deaths in total, when one sums deaths from different causes. Calculated per day, we get 1667 deaths per day in Italy on average, while the graph shows about 500 deaths as a baseline. It would be critical to know which regions or cities are included in the graph and which regions or cities are excluded, and the code used to produce the graph should ideally be published. --Dan Polansky (talk) 06:04, 8 April 2020 (UTC)


 * Dan Polansky, I was able to include population data, so I can plot per-capita numbers, and I also have a heat map showing the geographic coverage of the dataset. Regarding the Euro Momo comparison, I am not familiar with that dataset; just looking at the plots there, I can see a substantial peak in 2020 as well, albeit not as tall as in the Istat dataset. The Istat dataset is biased towards the most heavily infected parts of the countries, so that may explain the discrepancy. Regardless of which dataset is used, I think it would be important to show graphically the exceptional impact of the COVID-19 epidemic on total mortality. The code used to produce the plots is at https://github.com/allaisandrea/mortalita-covid. Allais.andrea (talk) 22:11, 12 April 2020 (UTC)

Updated plots with data published 2020-04-16. Dataset now extends to 2020-04-04, covers 1689 cities and 32% of the national population. Allais.andrea (talk) 15:18, 17 April 2020 (UTC)
 * @Allais.andrea: Let me tell you that the work you have done, with the updated graph and the map showing the dataset distribution, is very interesting. I am no epidemiologist, but I think the all-cause death statistics are going to prove more reliably indicative of the covid impact than anything else. --Dan Polansky (talk) 16:03, 17 April 2020 (UTC)
 * @Allais.andrea: Would you be able to provide the totals underneath the curves for the years? That is to say, would you be able to provide the total deaths from Jan 1 to Apr 16 for years 2015, ..., 2020? --Dan Polansky (talk) 10:08, 19 April 2020 (UTC)
 * Dan Polansky I am reluctant to include absolute death counts. I did that at the beginning of this conversation, and it caused a lot of confusion and even cast doubts on the reliability of the data. Maybe I'll do that if/when it will be possible to safely extrapolate the counts to the whole population. Allais.andrea (talk) 15:38, 20 April 2020 (UTC)
 * @Allais.andrea: Frankly, if the numbers cast doubt, it is even more important to have them. I think the doubt should be dispelled once we take into account the percentage of the population covered by the data set, which is indicated above as 32%. To remove any confusion: I do not ask you to redraw the graph; I ask for six cumulative figures, one per year. --Dan Polansky (talk) 16:32, 20 April 2020 (UTC)

Please consider adding to section 7.3 (Statistics.Charts) the following (Allais.andrea (talk) 17:14, 25 April 2020 (UTC)):




 * Red information icon with gradient background.svg Not done for now: please establish a consensus for this alteration before using the template. Aasim 22:59, 25 April 2020 (UTC)

Attempting to build consensus. Are there objections to the inclusion of these two graphs in the page? If so, what needs to change? Allais.andrea (talk) 01:36, 26 April 2020 (UTC)
 * I support the edit request. The source data for the graphs is known and identified and the code making the graphs was published. As a corroboration, the death count graph approximately matches the graph for Italy in ft.com. --Dan Polansky (talk) 11:44, 1 May 2020 (UTC)

Connection to Munich/Initial patient
I've adjusted the about the origin in Germany a bit. It was obviously assumed/reported that the outbreak was linked to the Munich cluster, so I let this part in. However, two of the three sources did not actually support that connection, so I commented the out. You also simply cannot use nextstrain (by itself) to make statements about how the virus propagated from one place to another. Unfortunately the Repubblica source is behind a paywall, so I cannot check the details.

For a scientific point, it appears more likely that the Italian outbreak is directly related to Wuhan than to Munich (I only found a German source, though the paper that was mentioned there was not directly quoted: )

Also the next sentence ("The 38-year old was asymptomatic for weeks") is presented as a fact - although it is only based on a single interview response, which itself was not based on any hard evidence (and it appears somewhat unlikely in the light of what is actually known about the disease). I think the statement should at least be qualified (or edited, once more information becomes available). Averell (talk) 13:43, 2 May 2020 (UTC)


 * Interestingly, the reason for so many cases in Bergamo, Lombardy was from a soccer match. In Madrid, Castile, Spain, it was from a large Women's Int'l Day gathering. In Mulhouse, Alsace, France from a church gathering. In Heinsberg-Gangelt, Germany was in a winter carnival. And the UK's first 2 weeks of March was an experimental "herd immunity" theory before the country locked down later than most of Europe on the 23rd. The 38 year old male as Italy's index case was back from Bavaria, he never visited towards the Netherlands border in Germany with its larger cluster, although Bavaria had villages which were hotspots. Covid-19 spread to many countries from Dec 2019 (confirmed Jan 1, 2020)-Feb (22nd with the first quarantines in Italy and since March-April, easing of lockdowns throughout China started in Wuhan, Hubei) and it's now found in 95% of the world's countries affecting over 3 million people (1/3 in the USA-60% in the Northeast states-half in New York City alone) as of May 1st. 2605:E000:100D:C571:7D82:A683:E434:DB3D (talk) 17:39, 3 May 2020 (UTC)

Semi-protected edit request on 5 May 2020
I would like to add following link to the External Links