Talk:Mortality due to COVID-19

CFR MLE Graph
To User:Debatable_narrative: might be right and it is very kind of you to have added it but it is very technical (for the average reader) in terms of statistical inference ("case fatality-derived risk of death" is a non-trivial concept to grasp). I am afraid it will make it confusing for the average reader (it is/was definitely confusing for me). You need to consider that this is not an article addressed to professional statisticians. Most people will not understand this graph and will be confused specially when they compare it to the numbers in the table. I am sure you are a very talented and intelligent person and I understand your graph provides interesting information but the article must be kept at a reasonable level of complexity so that it is quickly / easily readable and understandable by a majority of people without risk of confusion. This is not about mathematical intelligence (I have no doubt you rate high on that) but more about interpersonal intelligence (i.e. in this case making it easy for "average people" to be able to quickly read/understand the article), thank you so much for your understanding! 81.154.24.81 (talk) 22:32, 12 May 2020 (UTC)

Too soon
This article seems a bit rushed and questionable to me. For instance, it seem questionable to state that "scientific studies" assess the mortality as between 0.6 % and 0.9 % based on just two studies. Which seems to clash with the mortality rates registered in Taiwan (1.59 %) or New Zealand (1.4 %), which eradicated the epidemic by tracking all asymptomatic individuals; with the many Lombard towns (Castiglione d'Adda being the most striking example, with 80+ deaths from a population of 4,646) that saw the death of 1 % or more of their entire population; with data from New York based on antibody tests and excess deaths, which points to 1.4 % mortality.

Finally, a statement such as "in countries with high rates of COVID-19 infection, patients might be admitted to the hospital due to unrelated illnesses, but they could also have COVID-19. When these patients succumb to death due to the initial condition, their deaths are usually attributed to COVID-19 although the actual contribution of COVID-19 to the death is minimal" sounds rather misleading, and fuel for the conspiracy theories claiming that the number of covid deaths in the USA and other countries is being inflated because... reasons. In all pandemics, excess deaths are attributed to the pandemic, independently of what were the pre-existing conditions of the victims (healthy or already severely sick). This is not specific of COVID-19; if anything, data on excess deaths show that COVID-related deaths are underreported compared to past pandemics. --2.36.89.194 (talk) 13:21, 25 May 2020 (UTC)

Speaking of Excess Deaths

 * I'm sure the data is available out there, but all I could find was this article reporting that the overall death rate as dropped since the pandemic and the reaction to it. From April 28:
 * https://issuesinsights.com/2020/04/28/if-coronavirus-threats-so-great-whys-u-s-overall-death-rate-down/
 * One explanation is that the shutdowns and other precautions have lowered the incidence of other causes of death, particularly traffic accidents. Another issue is that Covid-19 may be the go-to cause of death for anyone dying of respiratory problems irrespective of testing for Covid-19 and this reduces the rates for other causes. An important figure would be the increase in death rates from respiratory illness since the pandemic. I hope a better researcher than I can find these answers now that, for example, road travel is getting back to normal, and so on. —Blanchette (talk) 23:11, 13 July 2020 (UTC)

May CDC IFR estimate
IFR estimates from this CDC report of May 22 calculating an IFR of .26% should be added under Epidemiology. https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html — Preceding unsigned comment added by Caracoid (talk • contribs) 22:18, 6 June 2020 (UTC)

Statistical Analysis
The page itself is ok, but I would encourage to make the dataset shown consistent, in particular at the "Mortality by age" part. Right now different countries are compared (which is ok), but the time span in comparison is different (which is not ok). IF we compare different countries then the same time span should be used. I understand that daily updates are not easily possible for many reasons (input of data may not happen by every country at the same time for example), but the analysis should be at the same time. Because otherwise, the data can not be compared 1:1. If some countries have data only up to May, and others including July, then the curves will be different. I suggest that the dataset should be standardized more; even if we can not get every day, I would recommend agreeing to something like the first of the month, or perhaps every second week or so. That way wikipedia could be more accurate in analysis of the data (there is sooooooo much bias out there in general by various newspaper, so I hope wikipedia can be a more objective medium). 2A02:8388:1641:8380:68DE:C1D0:C270:7927 (talk) 09:25, 14 July 2020 (UTC)
 * For China, for example, there are no further reports, but for the most part, the volunteers doing the updating simply stopped editing Template:COVID-19 CFR by age and country. The table was being updated more frequently a few months back. You can update it yourself if you care deeply. Also, I do not see any practical benefit of aligning measurements in time, rather than getting the latest data available. Comparing crude CFRs between countries in which epidemic peaked at different times, when most of them still have a lot of people in ICUs is not very informative either way. – attomir (talk &#124; contribs) 16:54, 14 July 2020 (UTC)

Merge and redirect proposal
There is substantial overlap between the content in these three articles as well as contradictory statements and out-of-date content in the two breakout topics:


 * Coronavirus disease 2019 (30,000 daily average PVs, 260 edits in prior 30 days)
 * Prognosis of COVID-19 (245 average daily PVs, 9 edits in prior 30 days)
 * Mortality due to COVID-19 (1,400 average daily PVs, 10 edits in prior 30 days)

I've merged unique content from this topic into Coronavirus disease 2019 and propose that we redirect this topic there. Please consider joining this discussion. - Wikmoz (talk) 21:53, 20 July 2020 (UTC)