working on the age breakout graphs for next post, but y’all may be interested in a preview:
You can really see the vaccine effect here – vaccines available in December 2020, after which % of COVID deaths among 85+ falls, etc. I’ll do a line graph with that…
I like graphs of raw rate of death by age over time to show the effect of vaccination. You can see a drop in each age group as it became eligible for vaccination.
I went so far as to create a set of graphs where each graph had one age cohort, and compared vaccination rate to infection rate over time for my state. When healthy young adults weren’t yet eligible and older adults were, the different shapes of the graphs was really striking.
The problem is that, in subsequent waves, the death rate went back up again for those age groups.
Thus this:
The vaccines are reducing risk for the vaccinated population… but the base risk has fluctuated, independent of what vaccines are doing.
I may do a ratio of death rates between age groups, which will get rid of the base risk and simply reflect how relative risks have changed.
We doing COVID deaths or Excess Deaths? Or do we (actuaries) not know expected deaths by age group?
I have been mostly looking at my state’s data. My state’s reported covid deaths closely match estimates of excess deaths, and won’t include stuff like overdoses. Also, I’m not a life actuary. I don’t know what expected deaths should be. And besides, death is only one of the stats I’ve been following. I’ve also been following vaccination, infection, and hospitalization rates.
Here: excess deaths discussion by age –
Supposedly I posted this already a couple weeks ago. [according to the forum software… I am not going to check directly]
“Video unavailable.” Probably just my work PC restricting me.
Works fine for me. Probably your net nanny.
Well, I could see why they might bar YouTube videos.
hmmm.
Anyway, I usually look at excess mortality, but wanted to focus on COVID specifically to not have the confounders of drug overdoses, etc., killing more younger adults affecting my lovely “wave” patterns. Because the younger adult excess mortality is more of a constant thing than “waves”
Here, from the CDC dashboard: (choose Weekly Number of Deaths by Age and update dashboard… and I’m going to focus on just one age group)
so, starting around spring 2020, there has been excess mortality for this young adult age group… and it’s been pretty constant. Only about 10% of that is COVID (except the most recent spike, which increased the COVID %age).
You stopin Sept 2021. I would really love ke to know what happens in the future
Mortality for the group is up~20% but only a tenth of the increase is covid, or is half the increase covid?
My time machine is in the shop.
So, mortality went up about 20% for the group, of that 20%, 10% is COVID (except for the most recent spike).
So basically, mortality went up 2% for that group due to COVID. The remaining 18% increase is due to other stuff.
How to you get to the 10%?
https://jamanetwork.com/journals/jama/fullarticle/2774445
Among adults aged 25 to 44 years, 4535 COVID-19 deaths were recorded, accounting for 38% (95% CI, 32%-48%) of the measured excess mortality.
I certainly would not attribute the entire increase to COVID, but this was still early on, and I’d expect some amount of undercounting of COVID deaths.
There was excess mortality for the entire year [look at my graph above]. It didn’t stop in July, when COVID deaths had already subsided after that first wave.
FWIW, there’s not much undercounting of COVID deaths in that first wave now, because death certificates have been updated.
Early on, when there weren’t enough tests, there was likely a lot of under-counting. If there any reason to believe the current numbers are any less reliable than the numbers for any other cause of death?
(Except 'heart attack". That seems to be the catch-all for a natural death that they don’t bother to investigate before filling out the paperwork. Both my father’s and my husband’s uncle’s death certificates say “heart attack”, and one died of a pulmonary embolism and the other of gall stones. I believe that heart attacks are greatly over counted in the official stats.)
Though, I’m about to correct myself.
This is a suspicious pattern:
It’s just that first wave with the spike of heart attacks. And then that goes away entirely. I heard COVID could cause blood clots… which then could get lodged in a cardiac artery, say. So maybe there still is a little undercounting of COVID from that first wave.
@meep Not sure if I should post this here or in the non-pandemic mortality thread. (OK, why not both).
They seem to put a lot of the blame for younger deaths on COVID. I thought you showed that was not the case.
Most of the excess mortality for everybody in the U.S. came from COVID. If you focus only on those under age 45, it was not COVID.
They note that:
Disease was not the only factor. Homicides and drug overdose deaths rose last year, too, said Bryan Tysinger, a research assistant professor in health policy at the University of Southern California, who was not part of the BMJ study, suggesting that the pandemic’s effects on social structures were a large part of the trend.
FWIW, other countries definitely did not have the large increase in homicides that occurred in the U.S. I’m not sure about the drug overdose issue, but increasing rates of ODs as cause of death has been a problem since fentanyl hit the scene.
race/ethnicity differences
and geography
