Avik Roy compares flu and COVID-19 deaths in eye-opening thread on risk by age group

Avik Roy, President of the Foundation for Research on Equal Opportunity, has put together this long and eye-opening thread that crunches the numbers by age group to compare flu deaths to those from COVID-19.

In summary, protect the elderly and open the schools (long bar is good, short is bad):

New @FREOPP: There is no request I get more often than to compare the risk of fatality from #COVID19 vs. the flu by age bracket. So I’ve put one together, assuming 150,000 deaths from COVID vs. the 2007-17 average from influenza & pneumonia combined. https://t.co/jIgUu9EErY pic.twitter.com/dhGJSa79ER

— Avik Roy (@Avik) May 18, 2020

For those 65+, the numbers get scary:

Assuming 150k #COVID19 deaths (the IHME mid-range estimate today), children are at much lower risk of dying from COVID vs. influenza or pneumonia. For those over 25, the relative risk is higher of death from COVID. But the *absolute* risk doesn’t get high until age 65: pic.twitter.com/swUhzacUzY

— Avik Roy (@Avik) May 18, 2020

If you’re looking at the first chart I posted (the one with the blue bars), a longer bar is “good” (i.e., relatively low risk of COVID death vs. influenza or pneumonia). A shorter bar is “bad” (relatively higher risk of COVID death). But absolute risk is most important.

— Avik Roy (@Avik) May 18, 2020

But all this depends on how many actually die this year from COVID-19:

Let me emphasize that this is an estimate, based on 150k U.S. COVID-19 deaths. If there are 300k deaths, all of these relative risk numbers would get divided by 2 (e.g., kids aged 5-14 would be 3.4x more likely to die of flu or pneumonia, not 6.8x more).

— Avik Roy (@Avik) May 18, 2020

For those who prefer the chart the other way—COVID deaths / flu+pneumo—here it is. In this version, the longer bar = higher relative risk of COVID. Shorter bar = flu & pneumonia riskier. Methodology & data sources @FREOPP: https://t.co/Sv7zI8Ve4G pic.twitter.com/bC0LFib70J

— Avik Roy (@Avik) May 19, 2020

This is potentially good news as governors get ready to re-open schools:

Here’s another way to think about the very low risk to children. Sweden never closed its schools, and yet Sweden has one of the lowest pediatric infection rates in the industrialized world (and exactly one pediatric death from #COVID19). https://t.co/PMBFcztSa7

— Avik Roy (@Avik) May 19, 2020

3 weeks ago, I criticized Texas @GovAbbott for ignoring the data on pediatric risk as it relates to school closures. Today, he & his staff announced that child care & summer school could reopen as of June 1, albeit with limitations (max class size 10). https://t.co/z6pCKmAQZd

— Avik Roy (@Avik) May 19, 2020

This chart compares COVID-19 to other causes of death:

Another chart from the new @FREOPP article (https://t.co/MBz1S8U3gW), this one comparing #COVID19 deaths by age bracket to deaths from all other causes, like cancer, heart disease, diabetes, homicide, and unintentional injury. pic.twitter.com/8aLCJGGK99

— Avik Roy (@Avik) May 19, 2020

In that chart, the light blue boxes at the bottom estimate COVID deaths by age bracket if there are 150,000 total deaths in the U.S. The grey-shaded boxes indicate causes of death that are less severe than COVID by age bracket; white/non-shaded boxes = more severe. pic.twitter.com/warJWgJcbt

— Avik Roy (@Avik) May 19, 2020

Less dangerous than diabetes for the 15-24 age bracket:

To take the 15-24 age bracket as an example: We estimate 161 deaths from #COVID19 in the U.S. for this group, vs. 13,895 from unintentional injury; 5,723 suicides; 5,172 homicides; 1,431 cancer; 949 heart disease; 388 birth defects; 211 diabetes; 189 influenza & pneumonia.

— Avik Roy (@Avik) May 19, 2020

Among those ≥65, #COVID19 would be the 5th leading cause of death at 120,118; behind heart disease (507,118); cancer (422,927); chronic lower respiratory disease (131,002); and cerebrovascular disease eg stroke (121,630). More severe than Alzheimer's Disease (114,883).

— Avik Roy (@Avik) May 19, 2020