Random Thoughts

Thank goodness, I qualified for free shipping on an SOA webinar.


I was fixing a bug in one of my programs when a fly appeared on the screen. For a moment I wondered if it was a new feature in Access.

Son, that’s what we used to call a bug.

An elderly mother of someone I know recently died of cancer. But because they tested positive for COVID (which had nothing to do with anything) the death certificate came back saying the reason for death was COVID.
I wonder how much COVID death overreporting is going on.

How exactly can you ever know this?

Probably not as much as the overreporting of heart disease. That’s what they put on the death certificate if someone dies of natural causes, and they don’t bother to check what happened before filling out the certificate.

I have two relatives whose death certificates say “heart attack” who died of something unrelated to the heart.

If you get a new phone, you will generate an enormous number of security alerts in the process of setting it up.

If you slice up some fruit and put it in a long sandwich, does that make it sublime?

Not much

Our resident COVID mortality expert (@meep) focuses on excess mortality so fortunately the COVID reporting worry isn’t relevant.

Best answer I have: some, but not enough to really matter. (agree with JSM)

The obsessive in me would love to throw your acquaintance’s mother’s death out of the COVID pile. However, Like Dara mentions, CoD isn’t “exact” anyway. How many vehicle deaths are caused by a heart attack resulting in crash? Should they be accidental deaths or just heart disease?

I think the key is excess mortality is the best starting place (agree with NormalDan). Seeing how that extra mortality lines up with other factors tells you more. If all COVID deaths happened to cancer patients receiving chemo on their DoD, I’d check the infusion equipment. You do see high excess death counts at advanced ages, but you are also seeing high relative excess deaths at lower ages as well. Does COVID kill at 5.3x flu deaths or 5.4x flu deaths (made up numbers) isn’t easily answerable with certainty. Does COVID kill at between 3x to 7x (again made up numbers) flu, better able to answer that.

FWIW, I think we have a better handle on epidemiology of COVID than we do on economic impact of COVID restrictions.

Don’t make me drop “they can put up to 30 causes of death on the certificate” on y’all.

[it averages to about 3 per cert., though]

More to the point – there does seem to be very high attribution of COVID for old folks… but not for younger folks among their excess mortality. Here’s the graph that really made me go HMM:

If you want the background on how I put that together, you can go here:

1 Like

[it really ought to have error bars on it. I’ll redo that in the next few weeks]

Wow… That makes me go “hmmm”, too.

I was flippant above. We are all data professionals here, and we all know that data is messy. One guy you know had cancer and also had covid-19, and he died. Probably they both contributed to his dying on that day. Data professionals ought to understand that “cause of death” isn’t a clean data element. It’s often more than one thing. (I’m surprised the average is as high as 3, but not that it’s well over 1.) I’m tired of people trotting out an anecdote about how all the covid-19 deaths are made up, when I’ve been looking at those scary excess mortality graphs since April or May.

But MPF’s graph is really surprising. And i doubt the issue is that they are miscoding the ages of covid-19 deaths.

Here is my post where I get into number of causes of death on a death certificate:

The main thing that caused me to go HMMMM is that COVID has not been a major direct contributor to excess mortality for those under age 45.

So I want to know what’s contributing to excess mortality for younger adults. Maybe it’s suicides & drug overdoses, but I don’t have access to that data.

I don’t know about nationally, but at least in New Hampshire overdose deaths are down:

(Of course, NH is a small state, so that might not hold up nationally.)

(EDIT: Looks like NY overdose deaths have significantly worsened, though. More generally

does not paint a good picture.

And some analyses of early pandemic data don’t seem to indicate a significant increase in (all-age) suicide (though that could have changed as things have dragged on); see for example:

Though that would still leave the question of where the younger age mortality is coming from.


Here is the COVID mortality thread:

and here is the general mortality trend thread:

I find it odd y’all are putting this stuff in here.

Well, it’s where the discussion was. Plus it’s one of the few threads where anything is pretty much on-topic. :laughing:

True enough, but if y’all have something substantive to say, it would be nice if you added to the other threads…

1 Like