Talk:COVID-19/All-cause deaths

Removed Sections of Learning Resource
The following sections were removed --Bert Niehaus (discuss • contribs) 12:28, 11 October 2020 (UTC)
 * Precautionary Principle
 * explaination about protective measures that explain, how public health introduces measures to mitigate the risk very early for communicable diseases before people die.
 * Preventive measures are introduced in the health system and public awareness was raised to reduced epidemiogical connectivity, when infections could be avoided.
 * As a subsection of a infectious disease page, I do not understand why the epidemiological context is removed. Consideration of the discussion about the precautionary principle lead to additional explainations in the learning resource to explain why decision making in epidemiology cannot wait until aggregated mortality shows an impact. Also those explainations are removed.
 * I did not notice this section earlier; it seems to have been posted to the top of the page contrary to normal post conventions, according to which new posts come to the bottom. Anyway. Any removals that I made came with edit summaries providing rationale, I hope. The edit summaries should provide all the explanation, but let me take up at least one point from the above: "explaination about protective measures that explain, how public health introduces measures to mitigate the risk very early for communicable diseases before people die": This has nothing to do with all-cause deaths and therefore does not belong to COVID-19/All-cause deaths; it does belong somewhere else in COVID-19. Keeping chapters/pages focused is important for their utility to the reader. --Dan Polansky (discuss • contribs) 13:56, 14 October 2020 (UTC)
 * As for precautionary principle: a discussion of the principle below under thread left us with no answer to the key question relating to the precautionary principle in relation to lockdowns: "Would you be able to provide links to artifacts in which the lockdown proponents established that the lockdowns will not result in significant harm?" At the same time, all-cause death charts are subject to which precautionary principle does not apply anyway; the principle could apply to a page pondering various interventive actions and inactions, but all-cause death is no such page. The all-cause death page and subpages do not propose any action or inaction and do not discuss any actin or inaction; it shows certain descriptive facts about the world. --Dan Polansky (discuss • contribs) 14:00, 14 October 2020 (UTC)

Precautionary principle
I do not see what precautionary principle has to do with all-cause deaths, and I would be inclined to delete all references to it from the page.

On another note, if we follow Precautionary principle's one formulation, let's see what we get:
 * under the precautionary principle it is the responsibility of an activity-proponent to establish that the proposed activity will not (or is very unlikely to) result in significant harm.

The activity proposed and implemented in many countries are lockdowns, whereas avoidance of imposition of lockdown is not an activity. Following the above, it is the responsibility of lockdown proponents "to establish that the proposed activity [lockdowns] will not (or is very unlikely to) result in significant harm"; such has not been done, and therefore, lockdowns shall be avoided if we follow the precautionary principle as formulated above.
 * "Doing A", "Doing B" and "Doing nothing" are all decisions that have an impact and could have a significant harm. Are you proposing that "Doing nothing" is the less harmful decision/"activity" in the sense?
 * Lockdowns are not performed e.g. for the Spanish flu. Activity proposed and implemented in many countries are lockdowns. The vaccination of the majority of population protects also a minority on unvaccinated people. Assume lockdowns in many countries are implemented to reduce epidemiological connectivity between people to protect the global spreading of a communicable disease as collaborative effort.  --Bert Niehaus (discuss • contribs) 05:26, 31 August 2020 (UTC)

That said, I do not see why precautionary principle should be on this page and I do not see that the above reasoning belongs to this page, or otherwise I would have added the reasoning to the page. I would sooner think it belongs to COVID-19/Precautionary principle or the like. --Dan Polansky (discuss • contribs) 11:39, 30 August 2020 (UTC)
 * Lockdown reduces epidemiological connectivity between people. If people do not meet they cannot infect each other. Nobody claims that this will not have side effects on other areas. I guess it is difficult to estimate how much the COVID-19 mortality would increase if states would not have applied the lockdown. If patients e.g.are afraid to get a necessary cancer treatment and that increases the mortality then this has to be addressed as well. Every life matters. Totat number of mortality is unspecific to identify the appropriate risk mitigation strategies.
 * Is there any error in my reasoning above? If so, which part of the reasoning is incorrect? Does precautionary principle speak in favor of the lockdowns or against them?
 * Precautionary principle does not speak in favor or against. The precautionary principle tries to select the decision option, that creates the minimal harm. Minimal harm is dependent on how you assess harm in different aeras of society, economy, culture and of course the health. If the you want to mitigate harm for economy you would come to a different decision in comparision to a health centric decision. A balance decision making can be possible if a least a basic introduction into the decision making process from different angles is visible in the learning resource. All-cause death is relevant parameter, but decision making in epidemiology is more complex than just looking on All-cause death. --Bert Niehaus (discuss • contribs) 13:57, 11 October 2020 (UTC)
 * "Totat number of mortality is unspecific to identify the appropriate risk mitigation strategies": Whether it is specific does not really matter that much as long as it is useful, by my assessment and per mortality.org if I understand correctly: "In response to the COVID-19 pandemic, the HMD team decided to establish a new data resource: Short-term Mortality Fluctuations (STMF) data series. Objective and internationally comparable data are crucial to determine the effectiveness of different strategies used to address epidemics. Weekly death counts provide the most objective and comparable way of assessing the scale of short-term mortality elevations across countries and time." --Dan Polansky (discuss • contribs) 14:59, 30 August 2020 (UTC)
 * You say "by my assessment". Do you want to publish your personal assessment here in Wikiversity if lockdowns are useful or not? --Bert Niehaus (discuss • contribs) 05:26, 31 August 2020 (UTC)
 * Aggregated mortality includes by definition all causes of death and does not show, if COVID-19 interventions decrease the number of deaths for cause A but increase the number of death for cause B. A scientific sound approach is to look into the details of causes - i.e. which cause of death decreased and which cause of death increased.
 * Furthermore lockdowns, social distancing, hygiene and self-protection are multiple measures to reduce the number of infections between people and they are introduced to flatten the curve, so that the health system can deal with the number of patients. Influenza vaccination is recommended to reduce the pressure on health system.
 * If you have doubts in that, what experts in public health and epdemiologist are proposing, then you should discuss with epidemiologist in scientific way with epidemiologist and public health experts.
 * If you agree, that epidemiologist and public health experts have the expertise and are trained in identification of appropriate responses and control measures for a epidemiological spread of an disease. If any of these interventions contributed to the total number of COVID-19 death, then you will not see these death in the all-cause deaths. Are you proposing to "do nothing and not reduce epidemiological connectivity" so that we can see increasing number in all-cause death in the mortality charts?
 * Fighting a pandemic is a collaborative effort of governments, public health agencies, the health system. Furthermore fighting a pandemic is especially dependent on the behaviour of the general public if a vaccination is not available.
 * Lockdowns, social distancing, hygiene and self-protection are measures to reduce epidemiological connectivity between people.
 * Vaccination of the population will prevent an epidemiological spread of a disease if a certain percentage of the population is vaccinated. Vaccination reduces the probability that an infected person can infect others, because a epidemic spread needs susceptible people that "meet" infected people (i.e. that they are epidemiologically connected). So a vaccination reduces the epidemiological connectivity between people by a pharmaceutical intervention. So the preventive measure is dependent on the majority of the population to be vaccinated. The same is applicable on non-pharmaceutical interventions it is a joint effort of multiple states.
 * Lockdowns, social distancing, hygiene and self-protection are non-pharmaceutical interventions that are introduced for COVID-19 are not virus specific interventions like a vaccination. A vaccination is designed for a specific virus. The non-pharmaceutical interventions are used, because we do not have a vaccination yet.
 * Precautional principle select the less harmful decision
 * --Bert Niehaus (discuss • contribs) 05:26, 31 August 2020 (UTC)
 * I am sticking to the subject, which is the analysis of the application of the precautionary principle as quoted; what measures I propose or oppose has no bearing to the analysis. The quoted precautionary principle is, again, this: "under the precautionary principle it is the responsibility of an activity-proponent to establish that the proposed activity will not (or is very unlikely to) result in significant harm": according to that principle, the burden to establish that the proposed activity, here lockdown, will not result in significant harm lies with the activity-proponent. Would you be able to provide links to artifacts in which the lockdown proponents established that the lockdowns will not result in significant harm? --Dan Polansky (discuss • contribs) 10:43, 2 September 2020 (UTC)
 * I appreciate that you take the effort of providing the analysis of all-cause deaths. It is a good source for learning. Epidemiological decision making is not as simple as looking at all-cause death curves and say lockdown is right or wrong choice in decision making. It is always a decision making under uncertainty and for a new virus the uncertainties are much greater than for a known virus. May be I should add a few explainations to learning resource, so that it is more comprehensive. --Bert Niehaus (discuss • contribs) 05:57, 3 September 2020 (UTC)
 * I don't appreciate that you add irrelevant text to resources that I try to create, here a link to "Precautionary principle", especially text that looks like your personal original research that traces to no sources; original research does not need to be bad, but we need to be clear that this is what it is, and I do have tracing to reliable sources analyzing all-cause death for covid. Therefore, I ask again: Would you be able to provide links to artifacts in which the lockdown proponents established that the lockdowns will not result in significant harm? Alternatively, can you please remove references to "Precautionary principle" from the page? --Dan Polansky (discuss • contribs) 13:36, 4 September 2020 (UTC)

Alleged additional noise
About "Aggregated data of all causes of death add additional noise to the COVID-19 data":

I find this misleading or even wrong. The covid data itself, namely covid-coded deaths, contains a lot of noise for the failure to distinguish covid-caused and covid-positive deaths; some countries even have official guidelines instructing to code a death as a covid one in case of doubt, casting doubt on the reliability of covid-coded death figures. I see no evidence or signs that covid-coded deaths are any less noisy than excess all-cause deaths. Yes, there is some noise in the excess deaths, but not necessarily more than in covid-coded deaths. --Dan Polansky (discuss • contribs) 11:32, 2 September 2020 (UTC)

There are similar problems with the following:
 * "if you want to analyze and assess the developement of a specific cause of death it is required to analyze the development of data for the specifc cause of death. Because scientifically you try to remove the noise from the data instead of adding noise to the data for COVID-19 directly."

I disagree with the above. The above is basically a dismissal of using all-cause deaths to get a more accurate picture of covid impact in terms of deaths. The dismissal does not trace to any source. --Dan Polansky (discuss • contribs) 11:42, 2 September 2020 (UTC)


 * If you want to analyze death caused by COVID-19, you must try to collect data about COVID-19 caused death. If you address the encoding you should refer international standardized mechanisms which must be international coordinated between member states and then implemented in the countries. Which takes time. If death of bicycle riders in traffic increase and at the same time the death of car drivers decrease, then you cannot conclude with the aggreated all-cause traffic death, that there is no problem for bicycle riders in traffic. Furthermore you cannot see in the aggregated data that protective measure in cars improved. A basic chart that combines both is a stackedarea chart with the total death cause and the fractions of specific death causes (see w:en:Template:Graph:Chart) --Bert Niehaus (discuss • contribs) 14:20, 11 October 2020 (UTC)


 * Years later: Reliable sources disagree with the above notion that to analyze death caused by COVID-19, we must look specifically at covid-coded deaths and not at elevations observed in all-cause deaths. Since: "Comparing the impact of the COVID-19 pandemic between countries or across time is difficult because the reported numbers of cases and deaths can be strongly affected by testing capacity and reporting policy. Excess mortality, defined as the increase in all-cause mortality relative to the expected mortality, is widely considered as a more objective indicator of the COVID-19 death toll.". There are limitations of using excess mortality for the purpose, such as inclusion of heat-wave-caused deaths in the excess mortality, but the limitations of covid-coded deaths are much graver for most countries, as the another article makes obvious: the article compares excess deaths and covid-coded deaths for many countries for years 2020-2021, showing a difference of factor 2 to be quite usual, in some cases even huger factor. Even industrial and meticulous Germany shows factor 1.82 betwen covid-coded deaths and excess deaths: 112 000 covid-coded deaths compared to 203 000 excess deaths in 2020-2021. This deficiency of covid-coding of deaths cannot be realistically solved for many countries across the world, and therefore, excess deaths calculated from all-cause deaths will remain a more reliable indicator of covid-related death toll for many countries in the years to come, and no realistic amount of increase of thoroughness of covid-testing will change that. --Dan Polansky (discuss • contribs) 05:47, 4 June 2022 (UTC)

Indoctrination vs. respect for the reader
The way I created this research resouce was very minimal. I provided charts of all-cause deaths, together with links to multiple reliable organizations supporting the notion that all-cause deaths are very relevant to covid analysis, and very little else. I kept the focus on the chart material, and I did not tell the reader what to think.

What was then added was what I can only describe as excessive biased indoctrination, treating the reader as a child who needs to be tutored by the state authority what to think. None of the excessive material added traces to reliable sources; in fact, it traces to no sources at all. The author of the indoctrination seems to think that he is above the linked sources and organizations.

This is very discouraging. It violates the requirement to trace to reliable sources and the requirement of neutrality.

I think the resource can still be useful for the reader if he ignores the indoctrination part and moves straight to the charts. --Dan Polansky (discuss • contribs) 09:36, 14 October 2020 (UTC)


 * It is good to trace to reliable source for mortality. No worries, I did not remove any your sources. I tried to explain, that mortality is in the time line at the end of the infection (preventive actions are before that that).
 * I added to the remark to look into additional peer-reviewed journals in epidemiology, virology. Public health interventions act before "all-cause death numbers" rise and interventions. You said, that you "created this research resouce" with very minimal effort. I would recommend to keep a remark on that page, that maybe also other peer-reviewed journals in Public Health, Virology, ... should be considered as well.
 * If you want so say something about the burdon of the health system, it makes sense to provide graphs for the capacity of the health system.
 * A learning objective of all-causes of death aggregates all-causes of death. So if that is learning objective. Learning task would be, what are patterns in the curve that can be observed.
 * If you want to present your research in Wikiversity. There is also the WikiJournal of Medicine with a reviewing process. --Bert Niehaus (discuss • contribs) 10:50, 14 October 2020 (UTC)
 * So far it seems that it is above all you who is presenting your original research. What are the sources for the material you have added? --Dan Polansky (discuss • contribs) 11:30, 14 October 2020 (UTC)
 * It seems you have added boldface to "your research". The subject of the thread is the indoctrination that you have added, one that traces to no sources at all. If you feel I added something that looks to you like "my research", one that in your view does not properly trace to sources or is problematic in any other way, we can discuss that, but please create a dedicated thread for that rather than changing the subject. --Dan Polansky (discuss • contribs) 13:16, 14 October 2020 (UTC)
 * I added a remark, to look also in peer reviewed jounrals virology, public health journals.--Bert Niehaus (discuss • contribs) 12:07, 14 October 2020 (UTC)
 * I am still waiting for at least one specific source, properly identified, for any of the material you have added. During the past months I was working on all-cause death pages, from what I remember, I saw zero provision of sources from you. It looks like I am supposed to take the statements you add to the pages on your authority, or on my independent review. --Dan Polansky (discuss • contribs) 13:05, 14 October 2020 (UTC)
 * Look on first comment on COVID-19 root page "Peer-reviewed Journals as evidence". Please feel free to remove all the comments that are not in line with your independent view. If you think that adding other peer-reviewed journals from virology, public health, ... is irrelevant for you, ignore my comment. All the best for your research, --Bert Niehaus (discuss • contribs) 13:32, 14 October 2020 (UTC)
 * Please help me and provide one specific source--not location pointing to a multitude of sources--supporting at least one specific statement that you have added to the page. That is how source traceability works. --Dan Polansky (discuss • contribs) 13:46, 14 October 2020 (UTC)

Sync w Wikidata?
It might be best if these data and visuals were drawn from Wikidata, and any data sourced for this purpose ended up there first -- that would help unify this work with that of the C19 wikiprojects on WP and WD. Is that already happening? –SJ + 03:05, 27 December 2020 (UTC)

Estimated excess mortality rate in Lancet 2022 - accuracy
I am perplexed about the estimated excess mortality rate in Lancet 2022.

For Czechia, Lancet 2022 indicates 49 100 excess deaths in 2020-2021, while it indicates excess deaths per 100,000 pop to be 244.8. By contrast, if we take Czechia population to be 10,524,167 per Czech Republic (2021 census), we get 49 100 / 10,524,167 * 100,000 = 466.5, a considerable mismatch against 244.8.

For Germany: Lancet 2022 indicates 203 000 excess deaths in 2020-2021, and the rate of 120.5 per 100,000 pop; Germany has 83,190,556 (2022 estimate) population, and we obtain the calculated rate of 203 000 / 83,190,556 * 100,000 = 244, a near double of 120.5.

For Bulgaria, Lancet 2022 indicates 82 500 excess deaths in 2020-2021, and the rate of 647.3 per 100,000 pop; Bulgaria has 6,863,422 (February 2022 estimate) population, and we obtain the calculated rate of 82 500 / 6,863,422 * 100,000 = 1202, a near double of 647.3.

What is going on? --Dan Polansky (discuss • contribs) 09:58, 2 June 2022 (UTC)

Necessary look at hospital data
I am about to remove the following:
 * "Look at the peaks in mortality in 2020 compared to other years? Explain why is it necessary to collect the data from hospitals with the available medical equipments to treat patients for specific medical support for COVID-19 and not look at aggregated data of mortality to adjust prepardness for communicable and non-communicable disease, treatment if injuries, ... (e.g. analyze mortality that is caused of peak hospitalizations and the associated healthcare overload by a higher demand for treatment of respiratory disease that cannot be covered by health system.)"

It is inaccurate: it is not strictly necessary to look at hospital load data, merely preferable. Absent hospital load data, sharp increases in deaths are indicative of sharp increases of hospital load; of course, hospital data is preferable, but that does not make death data useless for the purpose. The learner cannot explain what is not true. Furthermore, the alleged explaining that the learner is supposed to do cannot make any use of the data presented in the learning resource; it makes no actual use of the substantive content of the learning resource. --Dan Polansky (discuss • contribs) 14:59, 3 June 2022 (UTC)

Encoding of COVID-19 Death
I am about to remove the following item:
 * (Encoding of COVID-19 Death) Look at the difference between covid-coded deaths and covid-positive deaths in the section below. Keep in mind that any infection is a burdon to the immune system, even if the patient does not show any symptons (asymptomatic). Patients that have other risk factors are more vulnerable than patients without any previous diseases. Counting COVID-19 requires international standardized reporting mechanisms to make them comparable between countries. For a new disease the standardization is dependent on the scientific knowledge about the disease, that is increasing over time.
 * Beside the administrative challenge of international comparable statistics, keep in mind that the primary objective is to prevent people from getting the disease, getting to hospital, die and appear in the all-cause death statistics.

Starting with the first sentence, there is no difference between covid-coded and covid-positive deaths. Looking further, the sentences have nothing to do with all-cause deaths. The learner learns nothing from the sentences that pertains to all-cause deaths; in particular, the item does not propose any action for the learner in relation to the charts presented in the learning resource. As a minor issue, the infection being a burden on the immune system has nothing to do with "Encoding of COVID-19 Death"; the paragraph is incoherent. All-cause deaths do not provide any covid-coded data; they provide all data; thus, even the subject title "Encoding of COVID-19 Death" is of unclear relevance unless to point out that all-cause deaths are better in that they do not suffer of covid-coding problems such as under-reporinting resulting from insufficient testing, which is suggested by multiple references including Lancet 2022. I see nothing of value from the above paragraph that can be salvaged for the all-cause death resource. --Dan Polansky (discuss • contribs) 18:47, 3 June 2022 (UTC)

Let me quote: "Comparing the impact of the COVID-19 pandemic between countries or across time is difficult because the reported numbers of cases and deaths can be strongly affected by testing capacity and reporting policy. Excess mortality, defined as the increase in all-cause mortality relative to the expected mortality, is widely considered as a more objective indicator of the COVID-19 death toll.". --Dan Polansky (discuss • contribs) 18:54, 3 June 2022 (UTC)

Scientific reason to look at specific disease data that is not aggregated
I am about to remove the following: The item tries to explain that all-cause deaths are not as suitable as covid-coded deaths for true covid death impact, which contradicts reliable sources: "Comparing the impact of the COVID-19 pandemic between countries or across time is difficult because the reported numbers of cases and deaths can be strongly affected by testing capacity and reporting policy. Excess mortality, defined as the increase in all-cause mortality relative to the expected mortality, is widely considered as a more objective indicator of the COVID-19 death toll.". Furthermore, nowhere does the learning source suggest that all-cause charts should lead to abandonment of covid treatment or of all covid interventions; to the contrary, by looking at relatively bad and the worst excess deaths across the world, we may see what kind of bad outcomes can be prevented by appropriate measures. --Dan Polansky (discuss • contribs) 05:56, 4 June 2022 (UTC)
 * (Aggregated mortality and specific disease data) Now we compare aggregated mortality and disease specific data.
 * Aggregated data (as the title of this learning resource suggests) add up all causes of death and not only the data of a single disease. Take a non-communicable disease (e.g. breast cancer and remove the mortality of breast cancer from all years in mortality chart of all cases and add the breast cancer data of mortality just in 2020. Do you see an elevation above the normal mortality? If you do not see any elevation, would that justify to stop medical treatment because you analyzed aggregated data? What is the scientific reason to look at specific disease data, that is not aggregated to follow the development of the disease and to access the impact of specific risk mitigation strategies.

Furthermore, as for "Now we compare aggregated mortality and disease specific data", we don't: the learning resource does not feature covid-coded death charts (unlike e.g. Wikipedia), and the learning resource does not perform any such comparison. A comparison can be found e.g. in this article, showing excess deaths from all-cause deaths to be superior. --Dan Polansky (discuss • contribs) 06:01, 4 June 2022 (UTC)