All Cause Deaths Across Regions and Age Groups

Yesterday I drilled down to specific 6 week period in the winter for a few midwestern states.

Winter Deaths in Middle Aged Midwesterns

Looking for a way to capture trends across years, states and age groups, I came up with the representations seen below. Data source is same as yesterday, Weekly Counts of Deaths by Jurisdiction and Age. I split up the years into roughly flu season (Nov-Feb) & non-flu season (Mar-Oct) to make year over year comparisons more meaningful. Coincidentally & an added bonus is that this grouping also basically aligns with when COVID 1st entered the scene in a meaningful way (March 2020) and when vaccines were becoming widespread (Mar 2021).

Spring-Fall Months (Mar-Oct)

The maps below show a given states increase or decrease in all-cause deaths relative to the previous year’s season. So, for example, looking at the map in the far upper left we see California experienced 3% less deaths among 45-64 year olds in Mar-Oct 2018 vs. Mar-Oct 2017. Moving to the map in the far upper right, we see California experienced 5 % more deaths among 45-64 year olds in Mar-Oct 2021 vs Mar-Oct 2020.

What story do we see from the above?

• Spring-Fall 2018 was generally an improvement in deaths over 2017 for most age groups in most states.

• Mild-moderate increases in deaths, generally across the board in 2019.

• Moderate to significant increases in deaths in 2020. Obvious explanation is this is primarily an impact of COVID.

• 2021 looks even worse than 2020 in most places for 45-64 year olds. The 65-74 year olds look similar, but a bit better. As we move to 75-84 year olds, we see improvement in deaths in many places, although they certainly did not bounce back to 2019 levels. The oldest, 85+, showed improvement over 2020 in nearly all geographies (again, not back to 2019 levels however). Interestingly, many the places that did the worst in 2021 relative to 2020 appear to be the places where 2020 was relatively mild (WA, OR, ME, VT).

Winter Months (Nov-Feb)

• Winter 2018-19 was an improvement in deaths over 2017-18 for the 85+ group. For the other age groups, we see a regional pattern where mountain & pacific northwest states looked to have had a bad winter, while most of the rest of the country appears mild.

• Winter of 2019-20 appears to be relatively mild, with a few exceptions, across the board. Interesting in that it was coming of a Spring-Fall of 2019 that showed increased deaths over previous year.

• Pretty much we see significant increases in deaths in winter 2020-21 (1st COVID winter) in all age groups and all states. The CA→TX line of states looks particularly bad.

• Winter 2021-22 looks even worse than 2020 in most places for 45-64 year olds & we see a distinct regional pattern here, with the 65-74 year olds looking very similar with a bit milder numbers. Much like in the Spring-Fall time frames, for the age groups above 75 we see improvement in deaths in most places, but did not bounce back to 2019 levels. Again we see evidence of places that did not do as poorly in 2020-21 winter may have fared worse in 2021-22 (e.g. OR, WA, ME, VT).

The above does not take into account population trends. In other words, if I show a 2% increase in deaths among a certain age group in a state in a certain year, it may be that the population of that age group in that state increased by 2% that year, so an adjusted figure would be 0. If someone can point me to population estimates by year, age group, and state, I’d be happy to include (there’s chance its hanging out on my laptop somewhere but I’ve forgotten 🤔😂). In any case, population generally does not move that dramatically from year to year, so the general trends and patterns we see above should hold up.