Thread by @E_R_Sepulveda

1/N: My previous 9-country #COVID19 age-based case analysis was "static" due to data limitations, etc.

In this thread I present a dynamic analysis of case rates for high-risk age groups since the beginning of the pandemic in #Italy, #Spain, #Germany, #Canada & #Chile

Thread...

2/N: Focus is on 60+ age groups because, due to higher CFRs, these account for 90-95% of all fatalities

To reduce fatalities in current AND future #COVID19 waves, it is critical to understand case dynamics & whether "real" and generalized and if so, what are possible drivers?

3/N: @AndreasShrugged beat me to presenting dynamic analysis for #Germany & started interesting discussion, which can now be generalized based on 4+ countries

F3 presents same data differently (distinct periods, excl <60), with same results - an increase in cases for 80+ group

4/N: F1 presents #Italy age-based cases rates & shows that while abs. case rates higher than Germany, the relation between age groups is similar, specially in later periods (80+ cases are 2-3 x higher than 60-79)

Is this increase "real" or statistical artifact (testing, etc.)?..

5/N: F2 is #Spain age-based cases over time and shows very similar development as Italy, including spike in 80+ cases in last periods

So a 3 EU country pattern - independent of abs. case rates & initial cases, the 80+ cases increase relatively & absolutely as #COVID19 spreads...

6/N: F4 shows age-based cases for #Canada, which is "behind" most of EU for #COVID19, hence fewer data periods

In spite of this and being on different continent the pattern is replicated hear as well!

Note initial 80+ cases relatively lower, but then spiked in 2nd & 3rd periods

7/N: F5 shows age-based cases for #Chile, which is further behind #COVID19 wave & hence only include two periods

While early, same pattern of increase in 80+ cases...

Chile is important for pattern validation: South America (summer propagation), rel. young population, etc...

8/N: This 5-country analysis indicates pattern may be "real" - i.e not statistical artifact. Testing regimes vary significantly...and while all have prioritized 65+ testing (which could explain lower 60-69 cases), I don't think any has given priority to 80+ ages more than 70-79.

9/N: While all 5 countries are higher-income #OECD, they are sufficiently different to suggest pattern may be general... All of this preliminary because more time & observations needed (I could not find consistent time-series age data for #France, #USA. #China or #Korea)...

10/N: Meantime, containment/mitigation decisions have to be made, incl. with prelim./incomplete data

Is pattern driven by virology (higher "attack rate" for 80+) and/or by institutions (elder homes, multi-gen contact, etc.)?

We cannot yet change former, but we can impact latter

11/N: One possible future research topic: understand role of elder home institutions, including across countries.

For example, what % of 80+ in each country are in elder homes? Are 80+ case rates higher in elder homes compared to elders living with family or independently?

12/12: What are other inst. factors? We know about multi-gen contact, but how is it related to where 80+ elders live? How about 80+ income/wealth & ability to self-isolate?

In the meantime, I will continue to monitor these 5 countries & add others if I can find good data...

END