The leading cause of death in the US on Apr 6 was Covid-19 | Hacker News Reader

The leading cause of death in the US on Apr 6 was Covid-19

public.flourish.studio

80 points9nGQluzmnq3M136 comments

>> Disclaimer: COVID-19 data is the number of ACTUAL US deaths since March 15th, 2020 as reported on Worldometer against the backdrop of the EXTRAPOLATED DAILY number of deaths for top 15 causes of death in the US based on the latest (2018) data from the CDC. This chart is not meant to represent statistical analysis of any kind, it is meant for visual purposes only to help raise public awareness of the exponentially increasing COVID-19 deaths in the US

This is a really awful chart. The number of deaths from Covid-19 is animated as they increase day-by-day, but the numbers of deaths from all other causes remain constant. These numbers (the blue bars) are stated to be the "EXTRAPOLATED DAILY numbers of deaths from the top 15 causes of death in the US" (other than Covid-19), but if they are "DAILY" numbers, why aren't they changing daily?

The last sentence in the disclaimer is at least honest about it: this graph has no real information content. I for one find this graph to only serve the purpose of sensationalising a situation that is already dramatic enough without the need of animations and special effects. Really. The only thing missing is epic music in the background. Something from Carmina Burana, perhaps.

The "extrapolated daily" seems to be a weird way of saying "on average". And I think it's reasonable enough to do, because I don't think this data exists at daily granularity, and even if it did the daily variation would not be particularly meaningful.

The chart's really only making a single point, which is that yesterday, more people in the US (probably) died of COVID than anything else. And the figures are likely to get worse before they get better.

Since the other causes of death remain static in this chart, its hard to tell if Covid-19 are new cases or "take away from other causes" (and by what percentage).

Would be interesting to see daily chart of Covid-19 deaths as % of total deaths and daily total deaths.

PS. Extra points - number of deaths due to lockdowns (suicides, unaccessible healthcare, etc)

One data point: in New York City, deaths at home have increased from about 20-25 per day to 245, averaged over a recent week.

https://gothamist.com/news/surge-number-new-yorkers-dying-ho...

That's... more misleading thsn useful by itself. Of course when people they die they'll do so at home when they can't be somewhere else.

Indeed, this is the key question, what are the number of excess deaths compared with this time last year. Are mortality statistics available to that sort of granularity and speed (e.g. "number of deaths registered last week") for any country or region in the world?

If you die of heart failure and happen to be covid infected (which is likely given how widespread it is) do you go down as a covid death rather than a heart failure?

> Are mortality statistics available to that sort of granularity and speed

To answer my own question, yes

https://www.gov.uk/government/statistics/weekly-all-cause-mo...

I love gov.uk

Week 9 commences March 2nd (I'm assuming the weeks are the same as ISO weeks)

  Week | Total dead(E+W) | Covid (UK) | Excess mortality
  8  Feb24-Mar1 | 10,841 | 0    | ...
  9  Mar 2-8    | 10,816 | 3    | no
  10 Mar 9-15   | 10,895 | 32   | no
  11 Mar 16-22  | 11,019 | 300  | no
  12 Mar 23-29  | 10,645 | 1073 | no
  13 Mar30-Apr5 | n/a    | 3965 | yes -- England overall and 65+. not in Wales/Scotland/NI. Specific England regions - London, South East, E+W Midlands, North West

Note the graph showing the recent increase: https://i.imgur.com/McBopJql.png

The next weekly report is out tomorrow, which should have the total death figures for week 13.

It looks like in week 12 though, 10% of deaths in the UK were put down to COVID, but the total increase on the year before wasn't that high.

Next week will be englightening

Covid death figures from wikipedia.

GovUK also have this, hospital admissions. Look at the massive fall over the last few weeks.

https://assets.publishing.service.gov.uk/government/uploads/...

Ah great find. I was looking for UK hospital admissions data.

That's pretty shocking. Pneumonia and respiratory has hardly moved (where is the wave???) yet there's a huge fall in emergency cardiac patients. That's worrying. It implies people having heart attacks are choosing not to go to hospital fast enough, even though they could.

I'm assuming that half those attendances aren't because of heart attacks, but are to do with routine follow up appointments?

Isn't this data for emergency admissions only?

True, "Emergency departments"

https://www.bmj.com/content/bmj/369/bmj.m1406.full.pdf

People are ignoring stroke symptoms and failing to ring 999 because they fear being a burden on the NHS in England duringthe covid-19 pandemic, the national clinical director for stroke has warned. Deb Lowe, consultant stroke physician at Wirral University Teaching Hospital, said that doctors across the country were seeing “quite striking reductions” in the number of people coming into hospital with symptoms of stroke. She said, “It appears that people aren’t seeking emergency help or going to hospital when they suspect a stroke, possibly due to fear of the virus or not wanting to be a burden on the NHS.”

The excess deaths caused by the lockdown are likely to be spread out over years. They will be deaths of despair from suicide, substance abuse, and chronic diseases (including depression). The epidemiologists and public health officials don't appear to be factoring those deaths into their models.

The long tail of deaths that have been subverted must also be taken into account, e.g. take those who would have otherwise passed in the near future to lung disease who succumb to the virus, they inflate this number, but next year's lung disease numbers will be proportionately lower.

This graph shows total mortality in NY state https://twitter.com/JDVance1/status/1247728338484985857?s=20

From the disclaimer: COVID-19 data is the number of ACTUAL US deaths since March 15th, 2020 as reported on Worldometer against the backdrop of the EXTRAPOLATED DAILY number of deaths for top 15 causes of death in the US based on the latest (2018) data from the CDC.

And crucially ignores the reduction in deaths from heart failure, road traffic, etc. at this time

If the objective is to put the risk in perspective, then comparing to normal times makes more sense. Our perceived risk of dying in a car crash is based on the times we are actually normally driving and not when we are quarantined at home.

Heart failure sure

but road traffic related deaths? Really? COVID-19 is not causing a decrease in road traffic deaths, the reaction to COVID-19 is. So, why would that data ever be relevant when talking about the disease?

Because that also needs to be factored in when trying to determine causes

If total deaths doesn't change, but road traffic deaths are down by 500, that means non-road-traffic deaths are up 500

You need a complete picture to draw conclusions

Take a look at https://ourworldindata.org/coronavirus

This chart in particular:

https://ourworldindata.org/grapher/daily-deaths-covid-19?cou...

It's showing a doubling of daily confirmed deaths since last week. This is a lagging statistic, of course, but it's one of the more firm ones.

The total confirmed deaths can be seen here:

https://ourworldindata.org/grapher/total-deaths-covid-19?cou...

To put this into perspective, we're about halfway to the 23,000 flu deaths in the US for the entire flu season this year. At the current rate, though, we'll be past that by next week.

For more perspective, the 1957 Asian Flu killed about 70,000 people in the US (1 - 2 million worldwide), and our population at that time was 177 million (about half of today).

According to this chart:

https://ourworldindata.org/grapher/covid-confirmed-daily-dea...

we may be just at the start of bending the curve.

We're a much larger country than some of the others shown in our group (add Italy, France, UK, Spain, and Germany, and we still have more people).

EDIT:

According to: https://www.cdc.gov/flu/pandemic-resources/1957-1958-pandemi...

The number of deaths for the 1957 Asian Flu in the US was 116,000, at a time when we had half the population of today.

the link to the data for the CDC does not seem to match the graph...perhaps im reading this wrong?

https://www.cdc.gov/nchs/fastats/deaths.htm

if we're mapping based on the 2018 numbers the CDC provide then 49925 per month or ~12k per week is the cancer rate...

Diabetes is around 1700, not 200.

The chart is showing deaths per day

CDC guidelines appear to ask medical professional, to over-report COVID as a cause of death, if there is reasonable suspicion, but not a definite diagnosis.

I am not sure if this right/wrong, and how different it is from guidelines relating, to say, flu as a cause of death.

> "... In cases where a definite diagnosis of COVID–19 cannot be made, but it is suspected or likely (e.g., the circumstances are compelling within a reasonable degree of certainty), it is acceptable to report COVID–19 on a death certificate as “probable” or “presumed.” In these instances, certifiers should use their best clinical judgement in determining if a COVID–19 infection was likely. However, please note that testing for COVID–19 should be conducted whenever possible. ..." [1]

As an hypothetical example, if a deceased was in a contact with covid-positive person, but the deceased was not tested for Covid, should the cause of death be noted as 'presumed covid-19'?

[1] https://www.cdc.gov/nchs/data/nvss/vsrg/vsrg03-508.pd

It would be interesting to see the same chart but limited to just deaths that occur in hospitals.

A version that removes "one week average" as the latest value: https://public.flourish.studio/visualisation/1852838

At least the Covid-19 deaths have a limit, whereas many of the others will remain constant for a much longer time. The sadness of this disease is for many countries it did not need to be so bad. I wonder if we as a country (US) will learn anything from this.

What is the limit for Covid-19 deaths? We don't know yet if we'll get a working vaccine or if immunity from antibodies lasts long enough. What we have at the moment are educated guesses. If Covid-19 becomes endemic, it could be a death cause that is always present like influenza.

This might be a bit morbid, but every death cause has a limit, the whole population..

> This chart is not meant to represent statistical analysis of any kind

Okay.

Longer comment: "I'm admitting that my chart is totally meaningless, but it sure does look scary!"

How do you know that this was the cause of death when all we know is that these people were tested positive on Covid19 before they died? Not denying that Covid19 is a thing and measures are necessary but I think this question is valid when reading such headlines.

A positive test is not even necessary to report them as Covid-19 deaths in the US.

> Should “COVID-19” be reported on the death certificate only with a confirmed test?

> COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death.

https://www.cdc.gov/nchs/data/nvss/coronavirus/Alert-2-New-I...

Thanks!

> A positive test is not even necessary to report them as Covid-19 deaths in the US.

But this applies to other forms of death. We count flu deaths in the same way, so the data is equally skewed.

> > > COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death.

People can only put it on the death certificate if they can make that assumption to the best of their knowledge or belief. It's not doctors guessing.

> We count flu deaths in the same way, so the data is equally skewed.

We actually don't, because flu is not routinely tested for the way that COVID-19 currently is among suspected cases.

I think the comment claims that both flu and corona deaths are assigned almost solely based on symptoms and medical history (eg. if someone comes in with a fever, develops a pneumonia and dies then they look at signs to tell apart the diseases - did the patient have runny nose and joint aches? yes? then flu. no? but had shortness of breath? corona.)

It has a dramatic and unmistakable final phase? Plus, Occam's Razor

All mortality cause numbers are estimates, and the method is clearly stated in the website.

The title of the submission is somewhat misleading (we can't really say that the leading cause of death on that day was COVID-19), but the information is the best we have and it's helpful to make sense of the current crisis.

What you're asking can only be done months from now, once we have numbers for all deaths and their official causes.

I would bet this number is an underestimate, not at overestimate.

There are probably more people dying in their homes of covid-19 without ever having received a test (and thus not being counted) than people with a severe covid-19 reaction (since its still very hard to get tested if you have mild symptoms) but dying from something else.

So something else is causing a huge spike in mortality rates?

Wow, the rapid increase is definitely alarming.

Side note: gun violence would average number 3 or 4, if not for the tireless lobbying from the NRA to block health impact research. https://www.gunviolencearchive.org/

What we’ve learned from this Corona scare is the best way to protect people is to incarcerate them in their own homes. Take guns and hard instruments from people. Isolate everyone into their own cell would reduce the murder rate to zero. All communication would go through the Internet. No chance for physical abuse any more. No more transmission of disease since we’ve shutdown direct human interactions. It would be the perfectly safe world. Of course agency gets thrown out the window.

Reply on news.ycombinator.com